Is this you? Claim this profile to correct it, add a bio and choose the work people see first.
Claim this profileWorks335 from public data
- Pyramid Vision Transformer: A Versatile Backbone for Dense Prediction without Convolutions5,324
The Pyramid Vision Transformer (PVT) is introduced, which overcomes the difficulties of porting Transformer to various dense prediction tasks and is validated through extensive experiments, showing that it boosts the performance of many downstream tasks, including object detection, instance and semantic segmentation.
- Intern VL: Scaling up Vision Foundation Models and Aligning for Generic Visual-Linguistic Tasks3,476
A large-scale vision-language foun-dation model (Intern VL), which scales up the vision foun-dation model to 6 billion parameters and progressively aligns it with the LLM, using web-scale image-text data from various sources is designed.
- PVT v2: Improved baselines with pyramid vision transformer2,673
This work improves the original Pyramid Vision Transformer (PVT v1) by adding three designs: a linear complexity attention layer, an overlapping patch embedding, and a convolutional feed-forward network to reduce the computational complexity of PVT v1 to linearity and provide significant improvements on fundamental vision tasks.
- InternVL3: Exploring Advanced Training and Test-Time Recipes for Open-Source Multimodal Models1,936
InternVL3 incorporates variable visual position encoding (V2PE) to support extended multimodal contexts, employs advanced post-training techniques such as supervised fine-tuning (SFT) and mixed preference optimization (MPO), and adopts test-time scaling strategies alongside an optimized training infrastructure.
- Expanding Performance Boundaries of Open-Source Multimodal Models with Model, Data, and Test-Time Scaling1,813
This work explores the relationship between model scaling and performance, systematically exploring the performance trends in vision encoders, language models, dataset sizes, and test-time configurations and hopes this model contributes to the open-source community by setting new standards for developing and applying multimodal AI systems.
- 1,348
- InternImage: Exploring Large-Scale Vision Foundation Models with Deformable Convolutions1,217
This work presents a new large-scale CNN-based foundation model, termed InternImage, which can obtain the gain from increasing parameters and training data like ViTs, and reduces the strict inductive bias of traditional CNNs and makes it possible to learn stronger and more robust patterns with large- scale parameters from massive datalike ViTs.
- Shape Robust Text Detection With Progressive Scale Expansion Network860
A novel Progressive Scale Expansion Network (PSENet) is proposed, which can precisely detect text instances with arbitrary shapes and is effective to split the close text instances, making it easier to use segmentation-based methods to detect arbitrary-shaped text instances.
- Efficient and Accurate Arbitrary-Shaped Text Detection With Pixel Aggregation Network632
This paper proposes an efficient and accurate arbitrary-shaped text detector, termed Pixel Aggregation Network (PAN), which is equipped with a low computational-cost segmentation head and a learnable post-processing.
- BEVFormer: Learning Bird’s-Eye-View Representation From LiDAR-Camera via Spatiotemporal Transformers593
This work presents a new framework termed BEVFormer, which learns unified BEV representations from multi-modality data with spatiotemporal transformers to support multiple autonomous driving perception tasks, achieving outstanding results across these tasks.
- TAM: Temporal Adaptive Module for Video Recognition388
A new temporal adaptive module (TAM) to generate video-specific temporal kernels based on its own feature map that outperforms other temporal modeling methods consistently, and achieves the state-of-the-art performance under the similar complexity.
- Is Ego Status All You Need for Open-Loop End-to-End Autonomous Driving?296
This paper conducts thorough analyses and introduces a new metric to evaluate whether the predicted trajectories adhere to the road, and proposes a simple baseline able to achieve competitive results without relying on perception annotations.
- TEINet: Towards an Efficient Architecture for Video Recognition275
The proposed TEINet can achieve a good recognition accuracy on these datasets but still preserve a high efficiency, and is able to capture temporal structure flexibly and effectively, but also efficient for model inference.
- Efficient Deformable ConvNets: Rethinking Dynamic and Sparse Operator for Vision Applications267
The advancements in speed and efficiency of DCNv4, combined with its robust performance across diverse vision tasks, show its potential as a foundational building block for future vision models.
- DDP: Diffusion Model for Dense Visual Prediction254
The method, called DDP, efficiently extends the denoising diffusion process into the modern perception pipeline and shows attractive properties such as dynamic inference and uncertainty awareness, in contrast to previous single-step discriminative methods.
- GridFormer: Residual Dense Transformer with Grid Structure for Image Restoration in Adverse Weather Conditions240
The GridFormer framework achieves state-of-the-art results on five diverse image restoration tasks in adverse weather conditions, including image deraining, dehazing, deraining & dehazing, desnowing, and multi-weather restoration.
- Panoptic SegFormer: Delving Deeper into Panoptic Segmentation with Transformers199
This work presents Panoptic SegFormer, a general framework for panoptic segmentation with transformers, which contains three innovative components: an efficient deeply-supervised mask decoder, a query decoupling strategy, and an improved postprocessing method.
- Adaptive Graph Convolution for Point Cloud Analysis185
Compared with using a fixed/isotropic kernel, AdaptConv improves the flexibility of point cloud convolutions, effectively and precisely capturing the diverse relations between points from different semantic parts.
- 179
- PAN++: Towards Efficient and Accurate End-to-End Spotting of Arbitrarily-Shaped Text177
An end-to-end text spotting framework, termed PAN++, which can efficiently detect and recognize text of arbitrary shapes in natural scenes, based on the kernel representation that reformulates a text line as a text kernel surrounded by peripheral pixels.
- FB-BEV: BEV Representation from Forward-Backward View Transformations171
This work proposes a novel forward-backward view transformation module that compensates for the deficiencies in both existing methods, allowing them to enhance each other to obtain higher quality BEV representations mutually.
- Vision-RWKV: Efficient and Scalable Visual Perception with RWKV-Like Architectures166
Vision-RWKV (VRWKV), a model adapted from the RWKV model used in the NLP field with necessary modifications for vision tasks, surpasses ViT's performance in image classification and has significantly faster speeds and lower memory usage processing high-resolution inputs.
- Temporal Action Localization by Structured Maximal Sums155
This work addresses the problem of temporal action localization in videos by posing action localization as a structured prediction over arbitrary-length temporal windows, where each window is scored as the sum of frame-wise classification scores.
- Mini-InternVL: a flexible-transfer pocket multi-modal model with 5% parameters and 90% performance139
This work introduces Mini-InternVL, a series of MLLMs with parameters ranging from 1 billion to 4 billion, which achieves 90% of the performance with only 5% of the parameters, which makes these models more accessible and applicable in various real-world scenarios.
- VideoLLM: Modeling Video Sequence with Large Language Models134
This work proposes a novel framework called VideoLLM that leverages the sequence reasoning capabilities of pre-trained LLMs from natural language processing (NLP) for video sequence understanding and demonstrates that the understanding and Reasoning capabilities of LLMs can be effectively transferred to video understanding tasks.
- Graphology based handwritten character analysis for human behaviour identification113
An automatic method by analysing a few handwritten English lowercase characters from a to z to identify person behaviours by extracting structural features from different zones of isolated character images to derive the hypothesis based on a dictionary of Graphological rules.
- 109
- 97
- CNN‐RNN based method for license plate recognition96
This work explores the combination of Convolutional Neural Networks (CNN) and Recurrent Neural Networks namely, BLSTM (Bi-Directional Long Short Term Memory), for recognition, and proposes Dense Cluster based Voting (DCV), which separates foreground and background for successful classification of private and public vehicles.
- On Reinforcement Learning for Full-Length Game of StarCraft93
A hierarchical approach, where the hierarchy involves two levels of abstraction, which can reduce the action space in an order of magnitude yet remain effective and a curriculum transfer learning approach that trains the agent from the simplest opponent to harder ones.
- DCAN: Improving Temporal Action Detection via Dual Context Aggregation91
The Multi-Path Temporal Context Aggregation (MTCA) is design to achieve smooth context aggregation on boundary level and precise evaluation of boundaries and Coarse-to-fine Matching (CFM) is designed to aggregate context on the proposal level and refine the matching map from coarse to fine.
- CG-Bench: Clue-grounded Question Answering Benchmark for Long Video Understanding88
Cg-Bench is introduced, a novel benchmark designed for clue-grounded question answering in long videos and two novel clue-grounded white box and black box evaluations are designed, to assess whether the model generates answers based on the correct understanding of the video.
- Dynamic Sampling Networks for Efficient Action Recognition in Videos87
The experimental results demonstrate that DSN is able to greatly improve the inference efficiency by only using less than half of the clips, which can still obtain a slightly better or comparable recognition accuracy to the state-of-the-art approaches.
- MM-Interleaved: Interleaved Image-Text Generative Modeling via Multi-modal Feature Synchronizer83
MM-Interleaved is an end-to-end generative model for interleaved image-text data that introduces a multi-scale and multi-image feature synchronizer module, allowing direct access to fine-grained image features in the previous context during the generation process.
- 80
- Automatic analysis and integration of architectural drawings78
New methods to recognize typical structural objects and architectural symbols are proposed and the recognized results on the same floor and drawings of different floors will be integrated automatically for accurate 3D reconstruction.
- Eagle 2: Building Post-Training Data Strategies from Scratch for Frontier Vision-Language Models77
This work addresses VLM post-training from a data-centric perspective, showing the key role of data strategy in developing frontier VLMs and leads to a family of performant VLMs named Eagle2.
- Digital Twin of Intelligent Small Surface Defect Detection with Cyber-manufacturing Systems75
This article aims to construct cyber-manufacturing systems to achieve a DT solution for small surface defect detection task and shows the proposed algorithm could achieve high accuracy and recall in small defeat detection task, thus constructing DT in cyber- manufacturing.
- 74
- 72
- A new recognition model for electronic architectural drawings71
This paper proposes a Self-Incremental Axis-Net-based Hierarchical Recognition model for automatic recognition and interpretation of real-life complex electronic construction structural drawings and implements a series of integrated algorithms for recognizing dimensions, coordinate systems and structural components.
- Video Text Detection62
This book presents a systematic introduction to the latest developments in video text detection, covering pre-processing and post-processing techniques, character segmentation and recognition, identification of non-English scripts, techniques for multi-modal analysis and performance evaluation.
- 61
- OmniCorpus: A Unified Multimodal Corpus of 10 Billion-Level Images Interleaved with Text60
This paper introduces OmniCorpus, a 10 billion-scale image-text interleaved dataset that has 15 times larger scales while maintaining good data quality and is more flexible, easily degradable from an image-text interleaved format to pure text corpus and image-text pairs.
- Deep-dense Conditional Random Fields for Object Co-segmentation59
A deep-dense conditional random field framework is introduced to infer co-occurrence maps of an image by combining two parts: objectness scores of image regions and similarity evidences from object proposals across images.
- Refine-Net: Normal Refinement Neural Network for Noisy Point Clouds58
Qualitative and quantitative evaluations demonstrate the clear superiority of Refine-Net over the state-of-the-arts on both synthetic and real-scanned datasets.
- 58
- An automatic zone detection system for safe landing of UAVs58
An intelligent system for detecting regions to navigate a UAV when it requires an emergency landing due to technical causes is presented and Experimental results on images from different situations for safe landing detection show that the proposed system outperforms the existing systems.
- Mixed Link Networks57
This work presents a highly efficient and modularized Mixed Link Network (MixNet) which is equipped with flexible inner link and outer link modules and demonstrates that MixNets can achieve superior efficiency in parameter over the state-of-the-art architectures on many competitive datasets like CIFAR-10/100, SVHN and ImageNet.
- 55
- Spectral 3D mesh segmentation with a novel single segmentation field55
An automatic mesh segmentation framework that achieves 3D segmentation in two stages, hierarchical spectral analysis and isoline-based boundary detection, which is comparable to recent state-of-the-art algorithms.
- 49
- A novel character segmentation-reconstruction approach for license plate recognition48
Experimental results on benchmark license plate databases, as well as video databases, namely, ICDAR 2015, YVT video, and natural scene data, show that the proposed technique is effective and useful.
- A new multi-modal approach to bib number/text detection and recognition in Marathon images48
A new multi-modal approach to integrate torso and text detection approaches in a novel way to achieve good results in contrast to the conventional methods that rely on the characteristics of texts but not numerals of arbitrary orientations.
- Incremental Few-Shot Semantic Segmentation via Embedding Adaptive-Update and Hyper-class Representation47
This paper proposes an embedding adaptive-update strategy to avoid feature drift, which maintains old knowledge by hyper-class representation, and adaptively update category embeddings with a class-attention scheme to involve new classes learned in individual sessions to solve IFSS problems.
- 46
- 44
- 43
- Bayesian classifier for multi-oriented video text recognition system42
A new system to recognize video texts through binarization by introducing a Bayesian classifier and exploring wavelet decomposition and gradient sub-bands to enhance text information in video is developed.
- New Gradient-Spatial-Structural Features for video script identification40
This paper proposes to integrate the spatial and the structural features based on end points, intersection points, junction points and straightness of the skeleton of text components in a novel way to identify the scripts.
- Environmental sound classification for scene recognition using local discriminant bases and HMM40
This paper proposes to use the local discriminant bases (LDB) technique to identify the discriminatory time-frequency subspace for environmental sounds and then use it for corresponding feature extraction and the experimental results demonstrate the effectiveness of the proposed approach for auditory scene classification.
- 39
- Fractals based multi-oriented text detection system for recognition in mobile video images39
This work introduces fractals for text detection in video captured by mobile cameras in a novel way in the gradient domain and proposes to use k-means clustering for separating text components from non-text ones and direction guided boundary growing is proposed to extract multi-oriented texts.
- Docopilot: Improving Multimodal Models for Document-Level Understanding38
This work presents a high-quality document-level dataset, Doc-750K, designed to support in-depth understanding of multimodal documents, and develops a native multimodal model, Docopilot, which can accurately handle document-level dependencies without relying on RAG.
- 37
- Script independent approach for multi-oriented text detection in scene image37
Experimental results show that the proposed RRT for text detection in multi-oriented and multi-script environments outperforms the existing methods in terms of multi-lingual andMulti-oriented text detection ability.
- Fractional poisson enhancement model for text detection and recognition in video frames36
A generalized enhancement model based on fractional calculus to increase the quality of images obtained by Laplacian operation and is compared with standard enhancement models to show that the proposed model outperforms the existing models in terms of quality measures.
- Delaunay triangulation based text detection from multi-view images of natural scene34
The proposed method uses a recognition step to detect correct texts by comparing recognition results of view-1 and view-2, and the same recognition step is used for removing false positives to improve the performance of the proposed method.
- 34
- Rough-fuzzy based scene categorization for text detection and recognition in video33
Experimental results show that the proposed method outperforms the existing state-of-the-art methods, and the performances of text detection and recognition methods are also improved significantly due to categorization.
- 3D Reconstruction of Detailed Buildings from Architectural Drawings33
A new method for accurate 3D reconstruction from real-life architectural drawings is proposed that integrates and normalizes architectural information dispersed in multiple drawings and tables under the guidance of semantics and prior domain knowledge.
- 32
- Towards Ultra-Resolution Neural Style Transfer via Thumbnail Instance Normalization31
An extremely simple Ultra-Resolution Style Transfer framework, termed URST, to flexibly process arbitrary high-resolution images (e.g., 10000x10000 pixels) style transfer for the first time and surpasses existing SOTA methods on ultra-high resolution images benefiting from the effectiveness of the proposed stroke perceptual loss in enlarging the stroke size.
- 31
- Text segmentation in degraded historical document images31
A new combination of Sobel and Laplacian for enhancing degraded low contrast pixels is proposed for segmenting text and non-text in Indus documents based on the fact that text components are less cursive compared to non- text ones.
- A new DCT-PCM method for license plate number detection in drone images30
This work explores the combination of DCT and PCM (DCT-PCM) for detecting license plate number text in drone images and exploits fully connected neural networks for eliminating false positives to achieve better detection results.
- 29
- AE TextSpotter: Learning Visual and Linguistic Representation for Ambiguous Text Spotting28
This work proposes a novel text spotter, named Ambiguity Eliminating Text Spotter (AE TextSpotter), which learns both visual and linguistic features to significantly reduce ambiguity in text detection, and is the first time to improve text detection by using a language model.
- Data generation for connected and automated vehicle tests using deep learning models24
The results indicate that the generated data of the two generative models have reasonable differences while maintaining a certain similarity with the real samples, and the WGAN-GP model performs better than the VAE-GAN model according to the ratio of critical fragments.
- 24
- A new watershed model based system for character segmentation in degraded text lines23
Experimental results on different datasets show that the proposed method segments characters better than the existing character segmentation methods in terms of recall and precision.
- 23
- Integration of the Image-Guided Surgery Toolkit (IGSTK) into the Medical Imaging Interaction Toolkit (MITK)23
An integrated system developed by the integration of two open-source software (IGSTK and MITK) to shorten the development cycle of the image-guided surgery navigation system and save human resources simultaneously is presented.
- LLM-Align: Utilizing Large Language Models for Entity Alignment in Knowledge Graphs21
A Large Language Model (LLM) based Entity Alignment method, LLM-Align, which explores the instruction-following and zero-shot capabilities of Large Language Models to infer alignments of entities.
- A New Method for Detecting Altered Text in Document Images21
The proposed method explores the relationship between positive and negative coefficients of DCT to extract the effect of distortions caused by tampering by fusing reconstructed images of respective positive andnegative coefficients, which results in Positive-Negative DCT coefficients Fusion (PNDF).
- 21
- 20
- Graph attention network for detecting license plates in crowded street scenes20
Experimental results on the proposed Adaptive Progressive Scale Expansion based Graph Attention Network (APSEGAT) show that the method outperforms existing methods and is effective in detecting license plate numbers in crowded street scenes.
- 20
- Curved text detection in blurred/non-blurred video/scene images20
A new method for detecting texts in blurred/non-blurred images by estimating degree of blur in images based on contrast variations in neighbor pixels and a low pass filter, which results in candidate pixels for deblurring.
- Modeling spatial layout for scene image understanding via a novel multiscale sum-product network20
A new deep architecture MSPN is proposed for image segmentation, which utilizes multiscale unary potentials as the inputs and models the spatial layouts of image content in a hierarchical manner and characterizes scene spatial layouts in a fine-to-coarse manner to enforce the consistency in labeling.
- Efficient Reinforcement Learning for StarCraft by Abstract Forward Models and Transfer Learning19
The discovery that an abstract forward model [thought-game (TG)] combined with transfer learning is an effective way to accelerate reinforcement learning (RL) is presented and it is shown that the TG method is not as restrictive as it was thought to be.
- 19
- 19
- A new Fractal Series Expansion based enhancement model for license plate recognition18
A new Fractal Series Expansion (FSE) model for license plate image enhancement is presented and it is noted that recognition results after enhancement is higher than that of before enhancement, and hence the enhancement is useful and effective.
- 18
- 18
- Fractional means based method for multi-oriented keyword spotting in video/scene/license plate images17
A new model based on fractional means that considers neighboring information of pixels to widen the gap between text and background of input images is introduced and is effective and useful compared to the existing methods.
- 17
- A new unified method for detecting text from marathon runners and sports players in video (PR-D-19-01078R2)16
A comparative study with the state-of-the-art methods on bib number/text detection of different datasets shows that the proposed method outperforms the existing methods.
- 16
- Character shape restoration system through medial axis points in video16
This paper proposes histogram gradient division and reverse gradient orientation in a new way to select Text Pixel Candidates (TPC) for a given input character and proposes a novel way to restore the shape of the character from the PMAP.
- 16
- 16
- 16
- Multiclass object detection by combining local appearances and context16
A novel approach for multiclass object detection by combining local appearances and contextual constraints is presented, which achieves state-of-the-art performance in multiclass objects detection.
- A new ontology-based multimodal classification system for social media images of personality traits15
This paper focuses on classification of personality traits called Big Five factors, namely (i) agreeableness, (ii) conscientiousness, (iii) neuroticism, (iv) extraversion and (v) openness by combining image and textual features through ontology and fully connected neural network (FCNN).
- 15
- 15
- Forged text detection in video, scene, and document images14
A new Fourier spectrum-based method for detecting forged text in video images based on the premise that brightness distribution and the spectrum shape exhibit irregular patterns for forged text, while appearing more regular for original text.
- A New Forged Handwriting Detection Method Based on Fourier Spectral Density and Variation14
Experimental results show that the proposed method outperforms the existing methods in terms of classification rate.
- Learning discriminated and correlated patches for multi-view object detection using sparse coding14
This work proposes a novel method to perform multi-view object detection by learning discriminated and correlated patches firstly and then making inference based on them and the experiments performed illustrate the effectiveness of the proposed framework.
- 13
- 13
- Frequency Consistent Adaptation for Real World Super Resolution13
A novel Frequency Consistent Adaptation (FCA) that ensures the frequency domain consistency when applying existing Super-Resolution methods to the real scene and improves the performance of the SR model under real-world setting achieving state-of-the-art results with high fidelity and plausible perception.
- 13
- A New Deep Wavefront Based Model for Text Localization in 3D Video13
This paper proposes the first approach that successfully addresses the challenges of 3D video in addition to those of 2D, and overall outperforms existing state-of-the-art methods.
- A New U-Net Based License Plate Enhancement Model in Night and Day Images13
This paper presents a new work for enhancing license plates captured in limited or low light conditions such that license plate detection methods can be expanded to detect images at night.
- A New Multi-modal Technique for Bib Number/Text Detection in Natural Images13
A multi-modal technique that combines both biometric and textual features to achieve good results for bib number/text detection is presented and experimental results show that the proposed multi- modal technique outperforms the existing b Bib number detection method.
- 13
- A Hybrid Forecast Model of EEMD-CNN-ILSTM for Crude Oil Futures Price12
A hybrid forecast model of EEMD-CNN-ILSTM for crude oil futures price is proposed, which is based on Ensemble Empirical Mode Decomposition, Convolutional Neural Network, and Improved Long Short-Term Memory, and the results show the model is more effective and accurate.
- 12
- 12
- 12
- 12
- 12
- Content-oriented multimedia document understanding through cross-media correlation12
It is hypothesize that the correlation of different modalities from the same data source can help achieve better multimedia content understanding results than one which explores a single modality.
- Guiding Audio-Visual Question Answering with Collective Question Reasoning11
A novel Collective Question-Guided Network (CoQo), which accepts multiple question-answer pairs as input and leverages the reasoning over these questions to assist the model training process and can achieve state-of-the-art performance on three AVQA datasets while reducing training time significantly.
- How many trip requests could we support? An activity-travel based vehicle scheduling approach11
Three different versions of network representation and mathematical models for the activity-based vehicle routing problem to connect activity-travel graphs of passengers (demand layer) to spatio-temporal networks of vehicles (supply layer) are proposed.
- DCT-phase statistics for forged IMEI numbers and air ticket detection11
A new expert system for detecting forged IMEI numbers as well as altered air ticket images by derives the phase spectrum using the Discrete Cosine Transform (DCT) to highlight the suspicious regions and extracts phase statistics to study the effect of distortions introduced by forgery operations.
- 11
- 10
- MRSN: Multi-Relation Support Network for Video Action Detection10
The experiments demonstrate that modeling relations separately and performing relation-level interactions can achieve and outperformer state-of-the-art results on two challenging video datasets: AVA and UCF101-24.
- A new context-based feature for classification of emotions in photographs10
This work proposes the Context-based Features for Classification of Emotions in Photographs (CFCEP), which combines the Hanman transform values of foreground and background components according to their merits to generate two feature vectors that are fed to a CNN classifier for classification of images of different emotions uploaded on social media and public internet.
- A New Multiple-Distribution GAN Model to Solve Complexity in End-to-End Chromosome Karyotyping10
The proposed multidistributed generated advertising network (MD-GAN) to perform data enhancement by involving generated and sufficient training images and achieves high accuracy in tasks of chromosome detection, segmentation, and classification.
- Channel‐wise attention model‐based fire and rating level detection in video10
The authors propose a new method for predicting and rating fire in video through deep-learning models in this work such that rescue team can save lives of people.
- 10
- Visual Robotic Object Grasping Through Combining RGB-D Data and 3D Meshes10
A novel framework to drive automatic robotic grasp by matching camera captured RGB-D data with 3D meshes, on which prior knowledge for grasp is pre-defined for each object type.
- A new method for multi-oriented graphics-scene-3D text classification in video10
A novel method for classifying graphics-scene and 2D-3D texts in video based on the fact that static texts have zero velocity while dynamic texts have non-zero velocity to enhance text detection and recognition accuracies.
- A Novel Topic-Level Random Walk Framework for Scene Image Co-segmentation10
This paper puts forward a novel representation named Visual Relation Network (VRN) to organize multiple segments, and then search for meaningful segments for every image through voting on the network, and uses Scalable topic-level random walk to solve the voting problem.
- 9
- A new augmentation-based method for text detection in night and day license plate images9
This paper presents a new method for LPD based on augmentation and Gradient Vector Flow (GVF) in night and day images that uses a recognition concept to fix the bounding boxes, merging the bounded boxes and eliminating false positives, resulting in text/license plate detection in both day and night images.
- 9
- 9
- 9
- A Novel Multi-modal Integration and Propagation Model for Cross-Media Information Retrieval9
A novel Probabilistic Latent Semantic Analysis-based (PLSA-based) aspect model is presented and a new propagation pattern is proposed to refine multi-modal correlations by efficiently taking the complementary from multi- modalities.
- Saliency Regions for 3D Mesh Abstraction9
This paper proposes a3D mesh abstraction framework that is capable of segmenting a 3D mesh into specific parts called "saliency regions" based on a saliency computation process, by which the intrinsic pattern of the 3DMesh is promisingly able to be identified.
- 8
- 8
- New Fusion Based Enhancement for Text Detection in Night Video Footage8
A new fusion based enhancement method for text detection especially in night video footage that integrates the merits of color space and frequency based enhanced methods for sharpening low contrast details is proposed.
- Context-based environmental audio event recognition for scene understanding8
This is the first work that models event correlations as scene context for robust audio event detection from complex and noisy environments and average accuracy of audio event recognition can be effectively improved by capturing dominant audio sources and reasoning non-dominant events from the dominant ones through acoustic context modeling.
- Incremental 3D reconstruction using Bayesian learning8
A novel algorithm is presented, converting incremental 3D reconstruction to an optimization problem by combining two feature-enhancing geometric priors and one photometric consistency constraint under the Bayesian learning framework.
- 7
- 7
- Joint Alignment Networks For Few-Shot Website Fingerprinting Attack7
A novel few-shot WF attack called Joint Alignment Networks (JAN), which conducts fine-grained feature alignment at both semantic-level and feature-level, and outperforms the state-of-the-art few-shot WF methods, especially in the difficult 1-shot tasks.
- 7
- 7
- 7
- 7
- 7
- Deep invariant texture features for water image classification7
A variety of experimental results on a large dataset containing different types of clean and stagnant water images show that the proposed method outperforms the existing methods in terms of classification rate and accuracy.
- Rotation invariant angle-density based features for an ice image classification system7
A new method uses wavelet decomposition to extract robust features, such as those invariant to rotation, scaling and thickness of ice for classification, and is useful for accurate ice detection with better accuracy.
- 7
- 7
- 7
- Hand Pose Estimation with Attention-and-Sequence Network7
This paper presents a novel Attention-and-Sequence Network (ASNet) embedded with finger attention and joint sequence mechanisms, and achieves an average 3D error of 5.6 mm on ICVL, 10.3 mm on NYU, 7.3mm on MSRA, and these competitive results further confirm the great effectiveness of ASNet.
- 7
- 7
- 7
- A new ring radius transform-based thinning method for multi-oriented video characters7
Experimental results comparing with the state-of-the-art methods show that the proposed RRT to generate a radius map from Canny edges of each input image to obtain its medial axis is generic and outperforms the existing methods in terms of obtaining skeleton, preserving visual topology and recognition rate.
- 6
- Few-shot Semantic Segmentation via Perceptual Attention and Spatial Control6
A Diffusion-based Segmentation network (DiffSeg), which decouples probabilistic denoising and segmentation processes and equips it with a Spatial Control Module (SCM), which models spatial structural information of query images to control boundaries of attention maps, thus aligning the spatial location between knowledge representation and query images.
- 6
- Writer age estimation through handwriting6
A novel model for deriving the phase spectrum based on the Harmonic Wavelet Transform (HWT) for age classification on handwritten images from 11 to 65 years, which shows that the proposed method outperforms the existing methods in terms of classification rate.
- 6
- 6
- 6
- 6
- A Novel 3D Human Action Recognition Framework for Video Content Analysis6
This paper proposes to incorporate temporal patterns of joint positions with currently popular Long Short-Term Memory (LSTM) based learning to improve both accuracy and robustness in 3D human action analysis.
- 6
- 6
- 6
- 6
- A Dynamic-Rule-Based Framework of Engineering Drawing Recognition and Interpretation System6
This paper introduces the idea that recognition and interpretation of engineering drawings should be two interwoven phases, with each providing feedback to another, and applies this idea to a dynamic-rule-based method that was implemented in a system for recognizing and interpreting architectural structure drawings.
- 6
- 6
- Evaluation of declarative n-queens recursion: A deductive database approach6
The analysis shows that binding analysis and constraint exploration are two essential issues in the realization of declarative logic programming, based on investigation and experimentation using a deductive database approach.
- 5
- 5
- 5
- Highway On-Ramp Truck Platooning Based on Deep Reinforcement Learning5
A hierarchical merging strategy with a dynamic mechanism for selecting optimal merging points is introduced to facilitate efficient autonomous CAT platooning at highway on-ramps while considering the interference of HDVs.
- A new deep model for family and non-family photo identification5
A new method for the identification of photos of different types of families and non-families such that the method can assist investigation team to find a solution to human trafficking issue is presented.
- 5
- Structure Function Based Transform Features for Behavior-Oriented Social Media Image Classification5
A new method for the automatic classification of behavior-oriented images like Bullying, Threatening, Neuroticism-Depression, NeuroTicism-Sarcastic, Psychopath and Extraversion of a person from social media images is presented.
- A New Shadow Detection and Depth Removal Method for 3D Text Recognition in Scene Images5
This paper presents a new method for shadow detection and removal to enhance recognition performance for 3D texts in natural scene images using Gabor kernel with gray values of each image for separating shadow pixels from texts and background pixels.
- Symmetry features for license plate classification5
New symmetry features based on stroke width for classifying each input license image as private, taxi, cursive text, when they expand the symbols by writing and non-text such that an appropriate optical character recognition (OCR) can be chosen for enhancing recognition performance is presented.
- Context and Temporal Aware Attention Model for Flood Prediction5
This paper builds a context and temporal aware attention model for flood prediction based on a quantity of collected flood factors on top of Long Short-Term Memory (LSTM) networks, which selectively focuses on informative factors and pays different levels of attentions to the outputs of different cells.
- 5
- 5
- Text Detection in Multimodal Video Analysis5
This chapter introduces text detection in multimodal video analysis from a new view and introduces the relevance of different modalities existing in video, namely, the auditory, the visual, and the textual modalities.
- Introduction to Video Text Detection5
This chapter presents a brief historical overview to show how video text detection has evolved from the field of document image analysis and how the document analysis community has explored various methods proposed in different fields, including image processing, pattern recognition, computer vision, and artificial intelligence to find solution to text detection in video.
- A New Text Detection Algorithm for Content-Oriented Line Drawing Image Retrieval5
A new method for directly detecting texts from line drawing images is proposed, which is robust to text-graphic touching, scanning degradation and drawing noises, providing an efficient approach for content retrieval of document images.
- 5
- 5
- 5
- Specification and Modeling of Network Resources in Dynamic, Distributed Real-time System.5
The process of the invention is characterized by further comprising the steps of detecting the position of the movable trough adjacent to its critical or limit positions and adjusting the injection rate of the foaming gas so as to move the Movable trough to the middle position ofThe range of movement thereof.
- 4
- 4
- 4
- Classification of aesthetic natural scene images using statistical and semantic features4
A deep learning model is proposed in a novel way to fuse statistical and semantic features for the classification of aesthetic natural scene images by studying the naturalness of image content using statistical features, and reading text in the images using semantic features.
- 4
- 4
- Improved Ring Radius Transform-Based Reconstruction for Video Character Recognition4
This work proposes an Improved Ring Radius Transform (IRRT) for reconstructing impaired characters through medial axis prediction based on the Tangent Vector concept, which is effective, useful and outperforms existing methods.
- A deep action-oriented video image classification system for text detection and recognition4
A hybrid model for classification of action-oriented video images which reduces the complexity of the problem to improve text detection and recognition performance, and uses pixel, stable-region and face-component information in a noble way for solving complex classification problems.
- A New Context-Based Method for Restoring Occluded Text in Natural Scene Images4
A method for restoring occluded text to improve text recognition performance and uses the GOOGLE Vision API for obtaining labels for input images and the Natural Language Processing system known as BERT for extracting semantics between candidate words.
- Application of Back Propagation Neural Network Model in Prediction and Diagnosis of Osteoporosis4
The results of neural network analysis suggested that among the 12 factors influencing bone mineral density, gender had the most obvious effect, followed by height, weight and age, and the polymorphism of estrogen receptor PvuII had the greatest impact.
- A Spatial Density and Phase Angle Based Correlation for Multi-type Family Photo Identification4
A new method that combines spatial density information with phase angle for multi-type family photo classification is presented, which outperforms the existing methods in terms of classification rate for all the three databases.
- 4
- 4
- 4
- A Real-Time Mesh Animation Framework Using Kinect4
This paper proposes a real-time mesh animation framework to animate any kind of virtual characters by using Kinect, which is user-friendly and featured by real- time response for user motions and intuitive manipulations.
- 4
- 4
- 3
- 3
- 3
- 3
- Anomaly Handwritten Text Detection for Automatic Descriptive Answer Evaluation3
A novel method for detecting anomaly handwritten text in the responses written by the students to the questions based on the fact that when the students are confident in answering questions, the students usually write answers legibly and neatly while they are not confident, they write sloppy writing which may not be easy for the reader to understand.
- 3
- A New Hybrid Method for Caption and Scene Text Classification in Action Video Images3
Experimental results on five action classes and a comparative study with the existing methods demonstrate that the proposed method for the classification of caption and scene text in video images is effective.
- 3
- COMPRESSIVE SENSING BASED CONVOLUTIONAL NEURAL NETWORK FOR OBJECT DETECTION3
This paper proposes a method to compress convolutional neural network to decreases its computation and storage cost by exploiting inherent redundancy property of parameters in different kinds of layers of CNN architecture and evaluates the proposed method on several typical convolved neural network and shows it outperforms one latest compression approach.
- 3
- Efficient Reinforcement Learning with a Thought-Game for StarCraft3
The background knowledge to reinforcement learning in the form of a thought-game is incorporated and the trained agent can achieve a 100\% win-rate on the map \textit{Simple64} against the most difficult non-cheating built-in bot (level-7), and the training is 100 times faster than the previous ones under the same computational resource.
- TK-Text: Multi-shaped Scene Text Detection via Instance Segmentation3
This work presents a novel representation, i.e., text kernel, for multi-shaped texts, and proposes a simple yet effective scene text detection method, named as TK-Text, based on this representation.
- 3
- 3
- 3
- Cloud of Line Distribution and Random Forest Based Text Detection from Natural/Video Scene Images3
A new method based on Cloud of Line Distribution (COLD) and Random Forest Classifier for text detection in both natural and video images is presented to show its effectiveness and usefulness compared with the existing methods.
- Cloud of Line Distribution for Arbitrary Text Detection in Scene/Video/License Plate Images3
Experimental results and comparative study show that the proposed method outperforms the existing methods in terms of invariant to rotations, scripts and applications.
- Hierarchical Bayesian Network Based Incremental Model for Flood Prediction3
The proposed model not only appropriately embeds hydrology expert knowledge with Bayesian network for high rationality and robustness, but also designs an incremental learning scheme to improve the self-improving and adaptive ability of the proposed model.
- 3
- 3
- A New Multi-spectral Fusion Method for Degraded Video Text Frame Enhancement3
A new method that derives multi-spectral images from each input video frame by studying non-linear intensity values in Gray, R, G and B color spaces to increase the contrast of text pixels to enhance text information in degraded video frames is presented.
- Video Caption Detection3
This chapter focuses on the state-of-the-art methods developed for caption text detection in video, which make use of classifiers such as support vector machines, neural networks, and Bayesian classifiers.
- 3
- Movie Keyframe Retrieval Based on Cross-Media Correlation Detection and Context Model3
A novel cross-media correlation detection method for movie keyframe retrieval that gives improved retrieval performance in completeness and precision due to the efficient exploitation of the context and correlations between complementary multi-modalities.
- QuickDiagram: A System for Online Sketching and Understanding of Diagrams3
A system named QuickDiagram is proposed for quick diagram input and understanding that can recognize and beautify a component/symbol or a wire immediately and generation of PSpice codes from the recognized diagrams.
- A New Shape Descriptor for Object Recognition and Retrieval3
A new shape descriptor is presented for measuring the similarity between shapes and exploiting it in graphical object recognition and retrieval by statistically integrating the local length-ratio and angle constraints between contour points relative to the shape skeleton.
- 3
- 2
- 2
- 2
- YOLO-H: a lightweight object detection framework for helmet wearing detection2
This work proposed a lightweight real-time deep learning-based detection framework called YOLO-H, for automatic helmet wearing detection, which showed overwhelming performance in terms of speed and accuracy and can be applied to other object detection scenarios.
- 2
- 2
- A New Foreground-Background based Method for Behavior-Oriented Social Media Image Classification2
The proposed method introduces a multimodality concept using skin of different parts of human body and background information, such as indoor and outdoor environments, for the classification of behavior-oriented social images uploaded on various social media platforms.
- 2
- A New Common Points Detection Method for Classification of 2D and 3D Texts in Video/Scene Images2
The proposed method explores Generalized Gradient Vector Flow (GGVF) for finding dominant points for input 2D and 3D text line images based on opposite direction symmetry and recognition performance improves significantly as a result of applying it.
- SEE-LPR: A Semantic Segmentation Based End-to-End System for Unconstrained License Plate Detection and Recognition2
Comparative studies on several challenging benchmark datasets show that the proposed SEE-LPR system significantly outperforms the state-of-the-art systems in both accuracy and efficiency.
- 2
- Word-Wise Handwriting Based Gender Identification Using Multi-Gabor Response Fusion2
Experimental results on the dataset (multi scripts) apart from English, and benchmark databases, namely, IAM, KHATT, and QUWI, show that the proposed method outperforms the existing methods.
- 2
- 2
- HIRM: A handle-independent reduced model for incremental mesh editing2
This paper proposes a reduced deformation model that is independent of handle configuration, allowing users to dynamically update the handle set without noticeable waiting time, and supports incremental mesh editing where the user can freely define different handle sets to edit different parts of the shape without waiting for long re-computation.
- 2
- Weakly-supervised region annotation for understanding scene images2
A unified probabilistic graphical model called Topic-based Coherent Region Annotation (TCRA) for weakly-supervised scene region annotation that outperforms the recent state-of-the-art methods.
- Video Text Detection Systems2
This chapter introduces several typical real-life video text detection applications, including license plate recognition, navigation assistance, sport video analysis, and online video advertising.
- 2
- 2
- Character Segmentation and Recognition2
This chapter presents methods for character segmentation from text lines and recognition of video characters, and introduces a recognition method without the use of an OCR engine.
- Online stroke segmentation by quick penalty‐based dynamic programming2
A stroke segmentation method named quick penalty-based dynamic programming is proposed for splitting a sketchy stroke into several regular primitive shapes, such as line segments and elliptical arcs, which is robust for strokes with arbitrary shape and size.
- 2
- A Novel Approach for Robust Surveillance Video Content Abstraction2
A novel framework to efficiently and robustly convert a surveillance video clip into one abstraction image containing the integrated contour of interested objects is proposed, which has the following novelties: an improved w-SIFT algorithm for Y-axis frames offset calculation, a trapezoid-based compensation algorithm for X-axis perspective distortion correction, and an incremental video content integration approach.
- 2
- 2
- 3D Scene Analysis Using UIMA Framework2
A 3D scene analysis method by applying the UIMA (Unstructured Information Management Architecture) framework, which provides semantic-based intelligent annotation approaches and tools for unstructured data structures, to annotate VRML/X3D scene documents is proposed.
- 1
- 1
- 1
- Flowdashboard: authoring pandemic dashboards with a transparent flow model1
The proposed FlowDashboard framework accommodates all common components used in practical pandemic dashboards, and incorporate the pandemic gallery as an interface to facilitate users quickly learning the design space and demonstrate the usability and effectiveness of the framework.
- 1
- 1
- 1
- 1
- 1
- 1
- 1
- 1
- New Moments Based Fuzzy Similarity Measure for Text Detection in Distorted Social Media Images1
The similarity between text detection results given by three existing text detection methods is studied and the same similarity estimation is used in a novel way to remove extra background or non-texts and restoring missing text information.
- 1
- 1
- An Automatic System for Generating Artificial Fake Character Images1
The proposed system has the ability to generate fake and artificial scene characters that have similar shapes and colors with the existing ones and is validated by conducting text detection and recognition experiments on input and generated fake images to show its effectiveness.
- 1
- Robust Scene Text Detection for Multi-script Languages Using Deep Learning1
This work proposes a novel representation named Linked Extremal Regions (LER) to extract full characters instead of fragments of scene characters and proposes a two-stage convolution neural networks for discriminating multi-script texts in clutter background images for more robust text detection.
- A Novel Context-Aware Topic Model for Category Discovery in Natural Scenes1
This paper proposes a novel context-aware topic model for category discovery in complex natural scenes using a generative probabilistic procedure from three-level features consisting of patch, region and the entire image by introducing latent topic variables to every patch and every region.
- Fast GPU-based CT reconstruction applied in ablation treatment for hepatocellular carcinoma1
A visual assistance tool to assist the physician in providing patient-specific analysis and visualization based on graphic processing unit (GPU) hardware acceleration for clinical use in hepatocellular carcinoma (HCC) interventional planning is developed and represents an improvement in image guided preoperative planning.
- Retrieval of 3D CAD Models:Survey1
A comprehensive survey on the up-to-date retrieval methods for 3D CAD models was presented and existing difficulties and future research were discussed.
- Robust Shape Retrieval through a Novel Statistical Descriptor1
The proposed method's recognition rate for degraded shape images is greatly improved over a recent method (SFHM), and all the local distributions are integrated into a global distribution matrix for matching symbols.
- 3D Similarity Search Using a Weighted Structural Histogram Representation1
A fast and robust 3D retrieval method based on a novel weighted structural histogram representation that is effective and efficient for 3D shape retrieval and robust to translation, scaling or rotation transformations is proposed.
- –
- –
- –
- –
- –
- –
- A Brief Discussion on “Applying Reality to Reason” in Inorganic Chemistry under the Background of Big Data Digital Age–
The experimental teaching part of “Inorganic Chemistry” is also deepening the dynamic curriculum goal of “giving reality to reason” because of the enabling role of intelligent teaching.
- –
- –
- –
- –
- Revisiting Clear-Air Echo Classification in Cloudnet: A Deep Learning Perspective–
This study introduces and assesses a deep learning algorithm for identifying clear-air echoes across multiple sites and climatic conditions and highlights the robustness of DL methods and their potential to enhance cloud radar data quality in complex observational environments.
- –
- –
- –
- –
- A Knowledge Enforcement Network-Based Approach for Classifying a Photographer’s Images–
Experimental results of the proposed model show that the method outperforms the existing techniques by 5%–10% on average, and the dataset created for the experimental purpose will be made available upon publication.
- –
- Impact of Intelligent Equipment on Construction of Technical Regulations System of Beijing-Zhangjiakou High-Speed Railway–
The impact of various equipment on the existing operation process of high-speed railway and its adaptability to relevant technical regulations are summarized to provide reference for the revision of technical regulations of intelligent high-speed railway.
- –
- Context-Aware Residual Network with Promotion Gates for Single Image Super-Resolution–
A novel context-aware deep residual network with promotion gates, named as G-CASR network, for SISR, which offers insight for how to enhance contributive information and suppress useless information and requires only about 25% memory cost comparing with EDSR.
- –
- Multi-scale Comparison Network for Few-Shot Learning–
This paper presents a novel end-to-end model called Multi-scale Comparison Network (MSCN), which uses different scale convolutions for comparison to solve the problem of excessive gaps between target sizes in the images during few-shot learning.
- Text Component Reconstruction for Tracking in Video–
A new method for reconstructing text components especially from multi-views for tracking, which uses mosaicing concept for stitching PTC fromMulti-views based on overlapping regions between PTC, which results in reconstructed images.
- –
- Post-processing of Video Text Detection–
This chapter presents a method for character shape reconstruction using ring radius transform, a fusion method which combines wavelet and gradient bands for the text lines with the help of k-means clustering on every row and column to binarize the image.
- –
- Text Detection from Video Scenes–
This chapter gives an introduction to the current progress on scene text detection especially in the past several years, and starts from discussing the visual saliency of scene texts to describe the characteristics of text in natural scene images.
- A Parts-Based Multi-scale Method for Symbol Recognition–
The main idea is to decompose the symbol into the set of multi-scale local parts, some of which are not or less affected by the contextual interferences, and then recognize the symbol based on detecting and integrating individual symbol parts.
- Robust Object Tracking Using Motion Context in Crowded Scenes–
A novel framework to track motions for any object in a crowded scene even with occlusions is presented by using the modeled contexts of a tracking object through modeling its coupled motion context.
- –
- –
- –
- –
- COMPILATION AND EVALUATION OF NESTED LINEAR RECURSIONS: A DEDUCTIVE DATABASE APPROACH–
The analysis shows that nested linear recursions can be implemented systematically and efficiently using the chain-based query evaluation method.
Publication data from OpenAlex, with missing venues and authors filled in from Crossref; citation counts are the higher of OpenAlex and Semantic Scholar, last synced 2026-10-14. One-sentence summaries under some papers are written by Semantic Scholar’s model. Citation counts may be lower than on Google Scholar, which indexes more sources.
Report an error
Wrong papers, two people merged into one, or a profile that should not be here? Tell us and we will fix or hide it.