Picture for Hengcan Shi

Hengcan Shi

DrVideo: Document Retrieval Based Long Video Understanding

Add code
Jun 18, 2024
Figure 1 for DrVideo: Document Retrieval Based Long Video Understanding
Figure 2 for DrVideo: Document Retrieval Based Long Video Understanding
Figure 3 for DrVideo: Document Retrieval Based Long Video Understanding
Figure 4 for DrVideo: Document Retrieval Based Long Video Understanding
Viaarxiv icon

DifFUSER: Diffusion Model for Robust Multi-Sensor Fusion in 3D Object Detection and BEV Segmentation

Add code
Apr 06, 2024
Viaarxiv icon

JRDB-PanoTrack: An Open-world Panoptic Segmentation and Tracking Robotic Dataset in Crowded Human Environments

Add code
Apr 02, 2024
Viaarxiv icon

Open-Vocabulary Scene Text Recognition via Pseudo-Image Labeling and Margin Loss

Add code
Mar 12, 2024
Viaarxiv icon

Unified Open-Vocabulary Dense Visual Prediction

Add code
Jul 17, 2023
Viaarxiv icon

CoactSeg: Learning from Heterogeneous Data for New Multiple Sclerosis Lesion Segmentation

Add code
Jul 10, 2023
Viaarxiv icon

Open-Vocabulary Object Detection via Scene Graph Discovery

Add code
Jul 07, 2023
Viaarxiv icon

Class Enhancement Losses with Pseudo Labels for Zero-shot Semantic Segmentation

Add code
Jan 18, 2023
Viaarxiv icon

Transformer Scale Gate for Semantic Segmentation

Add code
May 14, 2022
Figure 1 for Transformer Scale Gate for Semantic Segmentation
Figure 2 for Transformer Scale Gate for Semantic Segmentation
Figure 3 for Transformer Scale Gate for Semantic Segmentation
Figure 4 for Transformer Scale Gate for Semantic Segmentation
Viaarxiv icon

ProposalCLIP: Unsupervised Open-Category Object Proposal Generation via Exploiting CLIP Cues

Add code
Jan 18, 2022
Figure 1 for ProposalCLIP: Unsupervised Open-Category Object Proposal Generation via Exploiting CLIP Cues
Figure 2 for ProposalCLIP: Unsupervised Open-Category Object Proposal Generation via Exploiting CLIP Cues
Figure 3 for ProposalCLIP: Unsupervised Open-Category Object Proposal Generation via Exploiting CLIP Cues
Figure 4 for ProposalCLIP: Unsupervised Open-Category Object Proposal Generation via Exploiting CLIP Cues
Viaarxiv icon