Is this you? Claim this profile to correct it, add a bio and choose the work people see first.
Claim this profileWorks157 from public data
- Domain Adaptive Faster R-CNN for Object Detection in the Wild1,634
This work builds on the recent state-of-the-art Faster R-CNN model, and design two domain adaptation components, on image level and instance level, to reduce the domain discrepancy, based on $$-divergence theory.
- Semantic Foggy Scene Understanding with Synthetic Data1,504
A complete pipeline to add synthetic fog to real, clear-weather images using incomplete depth information is developed and supervised learning with the authors' synthetic data significantly improves the performance of state-of-the-art CNN for SFSU on Foggy Driving.
- 1,069
- Continual Test-Time Domain Adaptation861
The proposed CoTTA proposes to stochastically restore a small part of the neurons to the source pre-trained weights during each iteration to help preserve source knowledge in the longterm and demonstrates the effectiveness of the approach on four classification tasks and a segmentation task for continual test-time adaptation.
- 778
- DAFormer: Improving Network Architectures and Training Strategies for Domain-Adaptive Semantic Segmentation650
A novel UDA method based on a Transformer encoder and a multi-level context-aware feature fusion decoder is proposed, DAFormer, which improves the state of the art in UDA by 10.8 mIoU for GTA→Cityscapes and 5.4 mIOU for Syruhia→City Scapes and enables learning even difficult classes such as train, bus, and truck well.
- MIC: Masked Image Consistency for Context-Enhanced Domain Adaptation412
A Masked Image Consistency module to enhance UDA by learning spatial context relations of the target domain as additional clues for robust visual recognition and significantly improves the state-of-the-art performance across the different recognition tasks for synthetic- to-real, day-to-nighttime, and clear-toadverse-weather UDA.
- Fast Optical Flow Using Dense Inverse Search393
The Dense Inverse Search-based method (DIS) is the efficient search of correspondences inspired by the inverse compositional image alignment proposed by Baker and Matthews (2001, 2004), making DIS ideal for real-time applications.
- 378
- End-to-End Urban Driving by Imitating a Reinforcement Learning Coach358
A reinforcement learning expert is trained that maps bird’s-eye view images to continuous low-level actions and provides in-formative supervision signals for imitation learning agents to learn from that achieves expert-level performance.
- Decoupling Zero-Shot Semantic Segmentation319
A simple and effective zero-shot semantic segmentation model, called ZegFormer, is proposed, which outperforms the previous methods on ZS3 standard benchmarks by large margins, e.g., 22 points on the PAS-CAL VOC and 3 points onThe COCO-Stuff in terms of mIoU for unseen classes.
- 316
- Dark Model Adaptation: Semantic Image Segmentation from Daytime to Nighttime305
A novel method to progressive adapt the semantic models trained on daytime scenes, along with large-scale annotations therein, to nighttime scenes via the bridge of twilight time, to alleviate the cost of human annotation for nighttime images by transferring knowledge from standard daytime conditions.
- MTR++: Multi-Agent Motion Prediction With Symmetric Scene Modeling and Guided Intention Querying297
The MTR framework achieves state-of-the-art performance on the highly-competitive motion prediction benchmarks, while the MTR++ framework surpasses its precursor, exhibiting enhanced performance and efficiency in predicting accurate multimodal future trajectories for multiple agents.
- Model Adaptation with Synthetic and Real Data for Semantic Dense Foggy Scene Understanding297
A novel method is proposed, named Curriculum Model Adaptation (CMAda), which gradually adapts a semantic segmentation model from light synthetic fog to dense real fog in multiple steps, using both synthetic and real foggy data.
- Deep Gradient Learning for Efficient Camouflaged Object Detection283
The proposed DGNet performs well in the polyp segmentation, defect detection, and transparent object segmentation tasks, and achieves comparable results to the cutting-edge model JCSOD-CVPR21 with only 6.82% parameters.
- 283
- Fog Simulation on Real LiDAR Point Clouds for 3D Object Detection in Adverse Weather251
This work addresses the challenging task of LiDAR-based 3D object detection in foggy weather by simulating physically accurate fog into clear-weather scenes, so that the abundant existing real datasets captured in clear weather can be repurposed for this task.
- Map-Guided Curriculum Domain Adaptation and Uncertainty-Aware Evaluation for Semantic Nighttime Image Segmentation212
A curriculum framework to gradually adapt semantic segmentation models from day to night through progressively darker times of day, exploiting cross-time-of-day correspondences between daytime images from a reference map and dark images to guide the label inference in the dark domains and a novel uncertainty-aware evaluation framework.
- LiDAR Snowfall Simulation for Robust 3D Object Detection196
This paper proposes a physically based method to simulate the effect of snowfall on real clear-weather LiDAR point clouds, and conducts an extensive evaluation using several state-of-the-art 3D object detection methods to show that this simulation consistently yields significant performance gains on the real snowy STF dataset.
- End-to-End Learning of Driving Models with Surround-View Cameras and Route Planners186
360-degree surround-view cameras help avoid failures made with a single front-view camera, in particular for city driving and intersection scenarios; and route planners help the driving task significantly, especially for steering angle prediction.
- Domain Adaptive Semantic Segmentation with Self-Supervised Depth Estimation174
This work proposes to explicitly learn the task feature correlation to strengthen the target semantic predictions with the help of target depth estimation, and uses the depth prediction discrepancy from source and target depth decoders to approximate the pixel-wise adaptation difficulty.
- Jointly Optimized Regressors for Image Super‐resolution166
A collection of regressors are jointly learned, which collectively yield the smallest super‐resolving error for all training data, and this method is conceptually simple and computationally efficient, yet very effective.
- Don’t Forget The Past: Recurrent Depth Estimation from Monocular Video162
This work is the first to successfully exploit recurrent networks for real-time self-supervised monocular depth estimation and completion and outperforms previous depth estimation methods of the three popular groups.
- Curriculum Model Adaptation with Synthetic and Real Data for Semantic Foggy Scene Understanding158
A novel method, named Curriculum Model Adaptation (CMAda), which gradually adapts a semantic segmentation model from light synthetic fog to dense real fog in multiple steps, using both labeled synthetic Foggy data and unlabeled real foggy data.
- Is image super-resolution helpful for other vision tasks?150
It is shown that applying ISR to input images of other vision systems does improve their performance when the input images are of low-resolution, and this work hopes this work will inspire the community to evaluate ISR methods also in real vision applications, and to adopt ISR as a pre-processing step of othervision tasks if the resolution of their input images is low.
- Scale-Aware Domain Adaptive Faster R-CNN142
This work presents Scale-aware Domain Adaptive Faster R-CNN, a model aiming at improving the cross-domain robustness of object detection by tackling the domain shift on two levels by explicitly incorporating the object scale into adversarial training.
- 139
- 129
- Weakly Supervised 3D Object Detection from Lidar Point Cloud127
This work proposes a weakly supervised approach for 3D object detection, only requiring a small set of weakly annotated scenes, associated with a few precisely labeled object instances, achieved by a two-stage architecture design.
- Three Ways to Improve Semantic Segmentation with Self-Supervised Depth Estimation124
This work proposes a framework for semi-supervised semantic segmentation, which is enhanced by self- supervised monocular depth estimation from unlabeled image sequences, and implements a strong data augmentation by blending images and labels using the geometry of the scene.
- 120
- Pix2NeRF: Unsupervised Conditional $\pi$-GAN for Single Image to Neural Radiance Fields Translation115
A pipeline to generate Neural Radiance Fields (NeRF) of an object or a scene of a specific class, conditioned on a single input image, based on $\pi$-GAN, a generative model for unconditional 3D-aware image synthesis, which maps random latent codes to radiance fields of a class of objects.
- Scribble-Supervised LiDAR Semantic Segmentation112
This paper proposes using scribbles to annotate LiDAR point clouds and releases ScribbleKITTI, the first scribble-annotated dataset for LiDar semantic segmentation, and presents a pipeline to reduce the performance gap that arises when using such weak annotations.
- Multi-Scale Interaction for Real-Time LiDAR Data Segmentation on an Embedded Platform105
The proposed Multi-scale Interaction Network (MINet) uses multiple paths with different scales and balances the computational resources between the scales and outperforms point-based, image- based, and projection-based methods in terms of accuracy, number of parameters, and runtime.
- 104
- Real-time and contactless measurements of thermal discomfort based on human poses for energy efficient control of buildings102
It is shown that individual thermal discomfort perception gives important feedback signals for energy efficient control of building heating, ventilation and air conditioning systems and there are few effective treatments for this problem.
- Depth Estimation from Monocular Images and Sparse Radar Data99
This paper gives a comprehensive study of the fusion between RGB images and Radar measurements from different aspects and proposed a working solution based on the observations, and demonstrates that the method outperforms existing fusion methods.
- Object Referring in Videos with Language and Human Gaze91
Experimental results show that the proposed novel network model for OR in videos, by integrating appearance, motion, gaze, and spatio-temporal context into one network outperforms previous OR methods.
- Learnable Online Graph Representations for 3D Multi-Object Tracking89
The approach provides a natural way for track initialization and handling of false positive detections, while significantly improving track stability, and employs a Neural Message Passing network for data association that is fully trainable.
- Talk2Nav: Long-Range Vision-and-Language Navigation with Dual Attention and Spatial Memory87
This work has developed an interactive visual navigation environment based on Google Street View, and introduces a soft dual attention mechanism defined over the segmented language instructions to jointly extract two partial instructions for matching the next upcoming visual landmark and the local directions to the next landmark.
- TrafficBots: Towards World Models for Autonomous Driving Simulation and Motion Prediction85
TrafficBots is presented, a multi-agent policy built upon motion prediction and end-to-end driving, and based on TrafficBots a world model tailored for the planning module of autonomous vehicles is obtained.
- PathTrack: Fast Trajectory Annotation with Path Supervision82
This work presents an efficient framework to annotate trajectories and uses it to produce a MOT dataset of unprecedented size, originally trained on the MOT15 dataset, and proves that this framework produces more accurate annotations than the state of the art, in a fraction of the time.
- Polarimetric SAR Image Classification Using Multifeatures Combination and Extremely Randomized Clustering Forests80
This paper firstly evaluates and compares different features for classifying polarimetric SAR imagery, then proposes two strategies for feature combination: manual selection according to heuristic rules and automatic combination based on a simple but efficient criterion.
- HGFormer: Hierarchical Grouping Transformer for Domain Generalized Semantic Segmentation79
Experiments show that HGFormer yields more robust semantic segmentation results than per-pixel classification methods and flat-grouping transformers, and outperforms previous methods significantly.
- Off-Policy Reinforcement Learning for Efficient and Effective GAN Architecture Search76
A new reinforcement learning (RL) based neural architecture search (NAS) methodology for effective and efficient generative adversarial network (GAN) architecture search that exploits an off-policy GAN architecture search algorithm that makes efficient use of the samples generated by previous policies.
- Both Style and Fog Matter: Cumulative Domain Adaptation for Semantic Foggy Scene Understanding72
A unified framework to disentangle the style factor and the fog factor separately, and then the dual-factor from images in different domains is devised, which achieves the state-of-the-art performance on three benchmarks and shows generalization ability in rainy and snowy scenes.
- CoSSL: Co-Learning of Representation and Classifier for Imbalanced Semi-Supervised Learning72
A novel co-learning framework (CoSSL), which decouples representation and classifier learning while coupling them closely and achieves state-of-the-art performance on benchmark datasets ranging from CIFAR-10, CIFar-100, ImageNet, to Food-101.
- Federated Incremental Semantic Segmentation71
A Forgetting-Balanced Learning (FBL) model is proposed to address heterogeneous forgetting on old classes from both intra-client and interclient aspects to address catastrophic forgetting in Federated learning-based semantic segmentation.
- Nighttime Defogging Using High-Low Frequency Decomposition and Grayscale-Color Networks70
This work addresses the problem of nighttime defogging from a single image by introducing a framework consisting of two modules: grayscale and color modules, which allows for consistency losses between the outputs of the two modules.
- 69
- Failure Prediction for Autonomous Driving66
The prediction method is able to improve the overall safety of an automated driving model by alerting the human driver timely, leading to better human-vehicle collaborative driving.
- 64
- HRFuser: A Multi-Resolution Sensor Fusion Architecture for 2D Object Detection61
The design of HRFuser is based on state-of-the-art high-resolution networks for image-only dense prediction and incorporates a novel multi-window cross-attention block as the means to perform fusion of multiple modalities at multiple resolutions.
- 59
- Improving Semi-Supervised and Domain-Adaptive Semantic Segmentation with Self-Supervised Depth Estimation56
A framework for semi-supervised and domain-adaptive semantic segmentation, which is enhanced by self-supervised monocular depth estimation (SDE) trained only on unlabeled image sequences, and achieves state-of-the-art results for semi-supervised semantic segmentation as well as for semi-supervised domain adaptation.
- Deep Domain Adaptation by Geodesic Distance Minimization55
This work builds on the recently proposed Deep CORAL method, which aims to train a convolutional neural network and simultaneously minimize the Euclidean distance of convariance matrices between the source and target domains, and proposes a new approach called Deep LogCORAL for unsupervised visual domain adaptation.
- Semantic Object Prediction and Spatial Sound Super-Resolution with Binaural Sounds53
This work develops an approach for dense semantic labelling of sound-making objects, purely based on binaural sounds, using a cross-modal distillation framework that consists of a vision `teacher' method and a sound `student' method to generate the same results as the teacher method.
- Domain Adaptive and Generalizable Network Architectures and Training Strategies for Semantic Image Segmentation51
HRDA, a multi-resolution framework for UDA&DG, that combines the strengths of small high-resolution crops to preserve fine segmentation details and large low- resolution crops to capture long-range context dependencies with a learned scale attention is proposed.
- 51
- 49
- Self-supervised Object Motion and Depth Estimation from Video46
A self-supervised learning framework to estimate the individual object motion and monocular depth from video as a 6 degree-of- freedom rigid-body transformation that eliminates the scale ambiguity of motion prediction through imposing a novel geometric constraint loss term.
- 46
- Benchmarking the Robustness of LiDAR Semantic Segmentation Models44
A robust LiDAR segmentation model (RLSeg) is designed which greatly boosts the robustness with simple but effective modifications and can boost future research in robust LiDAR semantic segmentation for safety-critical applications.
- 44
- Bi-level Alignment for Cross-Domain Crowd Counting41
This work designs a bi-level alignment framework (BLA) consisting of task-driven data alignment and fine-grained feature alignment, and introduces AutoML to search for an optimal transform on source, which well serves for the downstream task.
- Real-time 3D Traffic Cone Detection for Autonomous Driving40
This work investigates traffic cones, an object category crucial for traffic control in the context of autonomous vehicles, by exploiting the unique structure of traffic cones and proposing a pipelined approach to solve this problem.
- Cluster, Split, Fuse, and Update: Meta-Learning for Open Compound Domain Adaptive Semantic Segmentation39
This work proposes a principled meta-learning based approach to OCDA for semantic segmentation, MOCDA, by modeling the unlabeled target domain continuously, and achieves the state-of-the-art performance in both compound and open domains.
- 37
- 36
- 35
- 30
- 29
- 27
- 26
- 26
- 26
- 25
- 25
- 23
- 22
- 22
- 22
- 21
- 21
- 21
- 19
- 19
- 19
- 17
- 17
- 16
- 14
- 13
- 13
- 13
- 13
- 13
- 13
- 11
- 11
- 11
- 11
- 10
- 10
- 10
- 10
- 10
- 10
- 9
- 9
- 9
- Ret3D: Rethinking Object Relations for Efficient 3D Object Detection in Driving Scenes8
This work introduces a simple, efficient, and effective two-stage detector, termed as Ret3D, which utilization of novel intra-frame and inter-frame relation modules to capture the spatial and temporal relations accordingly.
- 7
- 7
- 7
- 7
- 6
- 6
- 5
- 5
- 5
- 4
- 4
- 4
- 4
- 4
- 4
- 3
- 3
- Guest Editorial: Special Issue on “Computer Vision for All Seasons: Adverse Weather and Lighting Conditions”3
The last decade has witnessed tremendous progress in many sub-fields of computer vision, due to the increasing availability of large annotated datasets such as PASCAL VOC, but the performance is mainly benchmarked under good weather and favorable lighting conditions.
- 3
- 2
- 2
- 2
- 2
- 2
- 2
- 2
- 2
- 2
- 1
- 1
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- High-level Feature Learning by Ensemble Projection for Image Classification with Limited Annotations–
- –
- –
- –
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 2026-10-13. 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.