Is this you? Claim this profile to correct it, add a bio and choose the work people see first.
Claim this profileWorks55 from public data
- LISA: Reasoning Segmentation via Large Language Model1,078
This work proposes a new segmentation task - reasoning segmentation and presents LISA: large Language Instructed Segmentation Assistant, which inherits the language generation capabilities of multimodal Large Language Models (LLMs) while also possessing the ability to produce segmentation masks.
- Mini-Gemini: Mining the Potential of Multi-Modality Vision Language Models415
This work introduces Mini-Gemini, a simple and effective framework enhancing multi-modality Vision Language Models (VLMs), and proposes to utilize an additional visual encoder for high-resolution refinement without increasing the visual token count to enhance visual tokens.
- Unifying Voxel-based Representation with Transformer for 3D Object Detection401
The proposed method achieves leading performance in the nuScenes test set for both object detection and the following object tracking task, and surpasses previous work in single- or multi-modality entries.
- Focal Sparse Convolutional Networks for 3D Object Detection326
This paper introduces two new modules to enhance the capability of Sparse CNNs, both are based on making feature sparsity learnable with position-wise importance prediction and shows that spatially learnable sparsity in sparse convolution is essential for sophisticated 3D object detection.
- Attention-Guided Unified Network for Panoptic Segmentation293
The underlying relationship between FG objects and BG contents is revealed, in particular, FG objects provide complementary cues to assist BG understanding, and the Attention-guided Unified Network (AUNet) is named, a unified framework with two branches for FG and BG segmentation simultaneously.
- Fully Convolutional Networks for Panoptic Segmentation202
This approach aims to represent and predict foreground things and background stuff in a unified fully convolutional pipeline and outperforms previous box-based and -free models with high efficiency on COCO, Cityscapes, and Mapillary Vistas datasets with single scale input.
- Learning Dynamic Routing for Semantic Segmentation190
A conceptually new method to alleviate the scale variance in semantic representation, named dynamic routing, which generates data-dependent routes, adapting to the scale distribution of each image, and compares with several static architectures, which can be modeled as special cases in the routing space.
- 162
- Voxel Field Fusion for 3D Object Detection121
This work presents a conceptually simple yet effective framework for cross-modality 3D object detection, named voxel field fusion, and develops mixed augmentor to align feature-variant transformations, which bridges the modality gap in data augmentation.
- Stitcher: Feedback-driven Data Provider for Object Detection76
Stitcher is presented, a feedback-driven data provider, which aims to train object detectors in a balanced way, and steadily improves performance by a large margin in all settings, especially for small objects, with nearly no additional computation in both training and testing stages.
- Multi-Scale Aligned Distillation for Low-Resolution Detection74
The idea of spatially aligning feature maps between models of varying input resolutions by shifting feature pyramid position is explored and crossed feature-level fusion is proposed to dynamically fuse teacher’s multi-resolution features to guide the student better.
- 66
- Fully Convolutional Networks for Panoptic Segmentation with Point-based Supervision65
A conceptually simple, strong, and efficient framework for fully- and weakly-supervised panoptic segmentation, called Panoptic FCN, which encodes each object instance or stuff category with the proposed kernel generator and produces the prediction by convolving the high-resolution feature directly.
- Development of an Immunochromatographic Strip Test for Rapid Detection of Melamine in Raw Milk, Milk Products and Animal Feed65
A simple, rapid and sensitive immunogold chromatographic strip test based on a monoclonal antibody for the detection of melamine (MEL) residues in raw milk, milk products and animal feed demonstrates that the developed method is suitable for the onsite determination of MEL residues in a large number of samples.
- State-Aware Re-Identification Feature for Multi-Target Multi-Camera Tracking42
In this framework, the occlusion status and orientation information are utilized in Re-ID model with human pose information considered and the tracklet association using the proposed fused tracking feature is adopted to handle the fragment problem.
- 39
- 27
- Vibrio harveyi co-infected with Cryptocaryon irritans to orange-spotted groupers Epinephelus coioides26
Findings provide the evidence of coinfections with C. irritans and V. harveyi in orange-spotted grouper.
- 25
- 21
- Attention-Aware Learning for Hyperparameter Prediction in Image Processing Pipelines21
This work proposes an attention-aware learning method that integrates the parameter prediction network into ISP tuning and utilizes the multi-attention mechanism to generate the attentive mapping be-tween the input RAW image and the parameter space.
- 21
- Variations of the Functional Brain Network Efficiency in a Young Clinical Sample within the Autism Spectrum: A fNIRS Investigation20
Results suggest individual differences of brain development in young children within the Autism spectrum, providing new insights into the psychopathology of ASD.
- 19
- An efficient and privacy-preserving truth discovery scheme in crowdsensing applications18
This work proposes an efficient and privacy-preserving truth discovery framework (EPTD-I) by adopting a novel data perturbation mechanism and improves the design with a user non-interactive scheme named EPTD-II to shift all encrypted truth discovery operations to cloud platforms.
- Attenuation of long-range temporal correlations of neuronal oscillations in young children with autism spectrum disorder16
The results suggest that the decreased DFA exponent of hemoglobin concentration signals may be one of the pathologic changes in ASD, and studying the temporal structure of brain activity via fNIRS technique may provide physiological indicators for autism.
- Novel analysis of fNIRS acquired dynamic hemoglobin concentrations: application in young children with autism spectrum disorder11
The results showed that there were information exchange deficits in the right prefrontal cortex of children with ASD, with markedly higher interregion SCFCs between the right PFC and other brain regions than those of normal controls, and the global SCFC was significantly higher in young patients with ASD.
- 11
- Production of virus-like particles of porcine circovirus 2 in baculovirus expression system and its application for antibody detection9
The findings suggest that this ELISA was rapid, specific, and reproducible and can be used for large-scale serological investigations of PCV-2 antibodies in pigs.
- 9
- Investigating the Validity and Reliability of a Comprehensive Essay Evaluation Model of Integrating Manual Feedback and Intelligent Assistance7
A fusion similarity algorithm for essay answer key points is proposed by extracting these characteristics that have an impact on the evaluation results, such as technique preferences, paragraph content and paragraph topic of the essays.
- 6
- An Evaluation Model of Online Autonomous English Learning Efficiency Using an Artificial Neural Network6
An evaluation model of online autonomous English learning efficiency based on artificial neural network is devised and the superiority of the improved model was demonstrated through experiments.
- Identity-Enhanced Network for Facial Expression Recognition6
A novel identity-enhanced network (IDEnNet) is presented to eliminate the negative impact of identity factor and focus on recognizing facial expressions, and achieves the best or comparable state-of-the-art on all three datasets.
- 5
- 5
- 5
- Simple and accurate approaches to predict the activity of benzothiadiazine derivatives as HCV inhibitors5
It proved that RBFNN and SVM were useful tools to predict the biological activity of benzothiadiazine derivatives as HCV inhibitors, and the proposed methods will be of significance in theHCV inhibitors research, and can be expected to apply to other similar research fields.
- 4
- 3
- Combined intrinsic decomposition and histogram matching for color migration3
The experimental results show that the new method proposed in the paper better preserves the texture structure of the result image, and the resulting result image has better visual effects and performs better in both subjective and objective evaluations.
- Extracting Process Graphs from Texts via Multi-Granularity Text Classification3
The experimental results demonstrate that the proposed hierarchical model to effectively model and extract multi-granularity information without manually defined procedural knowledge outperforms the state-of-the-art methods with statistical significance, and the ablation studies demonstrate its effectiveness.
- Complicated intermittent scratches detection research on surface of optical components based on adaptive sector scanning algorithm cascading mean variance threshold algorithm3
A novel intermittent scratch detection method based on adaptive sector scanning algorithm (ASC) cascading mean variance threshold algorithm (MVTH) with high degree of parallelism, high speed and strong robustness is brought forward.
- Adaptive grayscale method based on weighted guided filtering2
A new contrast retention grayscale method based on weighted guided filtering is proposed, and the saturation feature holding term is designed to better retain the contrast characteristics of the original color image.
- 1
- Defect detection method for complex surface based on human visual characteristics and feature extracting1
Experimental results show that the improved grayscale adjustment and frequency-tuned saliency algorithm combined with the salient region mask obtained by dilation and differential operation can retain defect information completely and achieve considerable extracting efficiency and detecting accuracy.
- 1
- 1
- Color transfer based on cartoon decomposition and efficient illuminant estimation–
The proposed color transfer method combining cartoon decomposition and illuminant estimation can not only effectively maintain the texture details of the source image but also achieve good results in terms of objective evaluation indexes, compared with the traditional color migration method, while achieving better visual effects.
- Contrast preserving decolorization method under gradient domain global mapping–
By studying the projection amplification of global mapping, a bimodal mapping method is designed by using gradient correlation as the standard and it is shown that under the premise of enhancing the gray contrast of the color image, the detail features of the original color image can be better preserved.
- –
- –
- A defect detection algorithm based on statistical feature of local visual field for complex metal curve surface–
To detect complex curved surface on semi-finished metal shell of cell phone, the experimental results demonstrate that the defects detection accuracy of the proposed algorithm can reach 95%, and the detection time for single test area is less than 1ms, which is suitable for accurate and real-time detection on the production line for such surface defect.
- –
- File layout brief introduction and mutually between change–
The file layout (File Formats) is one kind of the form a document is preserved in order to being unlike way being in progress, but, the image document is the document having described one image computer disc.
Publication data from OpenAlex; citation counts are the higher of OpenAlex and Semantic Scholar, last synced 2026-10-15. 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.