Shu Kong

papers

1,002

total citations

papers (18)

Domain Decluttering: Simplifying Images to Mitigate Synthetic-Real Domain Shift and Improve Depth Estimation

CVPR 2020arXiv

citations

OV-PARTS: Towards Open-Vocabulary Part Segmentation

NEURIPS 2023arXiv

citations

Few-Shot Recognition via Stage-Wise Retrieval-Augmented Finetuning

CVPR 2025arXiv

citations

Instance Tracking in 3D Scenes from Egocentric Videos

CVPR 2024arXiv

citations

Boosting Image Restoration via Priors from Pre-trained Models

CVPR 2024arXiv

citations

Continual Learning with Evolving Class Ontologies

NEURIPS 2022arXiv

citations

Camera Pose Matters: Improving Depth Prediction by Mitigating Pose Distribution Bias

CVPR 2021arXiv

citations

Solving Instance Detection from an Open-World Perspective

CVPR 2025arXiv

citations

CriSp: Leveraging Tread Depth Maps for Enhanced Crime-Scene Shoeprint Matching

ECCV 2024arXiv

citations

Information-Bottleneck Driven Binary Neural Network for Change Detection

ICCV 2025arXiv

citations

A High-Resolution Dataset for Instance Detection with Multi-View Object Capture

NEURIPS 2023

citations

Attention to the Burtiness in Visual Prompt Tuning!

ICCV 2025

citations

LCA-on-the-Line: Benchmarking Out of Distribution Generalization with Class Taxonomies

ICML 2024

citations

Shu Kong

papers (18)

OpenGAN: Open-Set Recognition via Open Data Generation

Long-Tailed Recognition via Weight Balancing

Multimodal Object Detection via Probabilistic Ensembling

Alpha-CLIP: A CLIP Model Focusing on Wherever You Want

The Neglected Tails in Vision-Language Models

Domain Decluttering: Simplifying Images to Mitigate Synthetic-Real Domain Shift and Improve Depth Estimation

OV-PARTS: Towards Open-Vocabulary Part Segmentation

Few-Shot Recognition via Stage-Wise Retrieval-Augmented Finetuning

Instance Tracking in 3D Scenes from Egocentric Videos

Boosting Image Restoration via Priors from Pre-trained Models

Continual Learning with Evolving Class Ontologies

Camera Pose Matters: Improving Depth Prediction by Mitigating Pose Distribution Bias

Solving Instance Detection from an Open-World Perspective

CriSp: Leveraging Tread Depth Maps for Enhanced Crime-Scene Shoeprint Matching

Information-Bottleneck Driven Binary Neural Network for Change Detection

A High-Resolution Dataset for Instance Detection with Multi-View Object Capture

Attention to the Burtiness in Visual Prompt Tuning!

LCA-on-the-Line: Benchmarking Out of Distribution Generalization with Class Taxonomies

papers (18)

OpenGAN: Open-Set Recognition via Open Data Generation

Long-Tailed Recognition via Weight Balancing

Multimodal Object Detection via Probabilistic Ensembling

Alpha-CLIP: A CLIP Model Focusing on Wherever You Want

The Neglected Tails in Vision-Language Models

Domain Decluttering: Simplifying Images to Mitigate Synthetic-Real Domain Shift and Improve Depth Estimation

OV-PARTS: Towards Open-Vocabulary Part Segmentation

Few-Shot Recognition via Stage-Wise Retrieval-Augmented Finetuning

Instance Tracking in 3D Scenes from Egocentric Videos

Boosting Image Restoration via Priors from Pre-trained Models

Continual Learning with Evolving Class Ontologies

Camera Pose Matters: Improving Depth Prediction by Mitigating Pose Distribution Bias

Solving Instance Detection from an Open-World Perspective

CriSp: Leveraging Tread Depth Maps for Enhanced Crime-Scene Shoeprint Matching

Information-Bottleneck Driven Binary Neural Network for Change Detection

A High-Resolution Dataset for Instance Detection with Multi-View Object Capture

Attention to the Burtiness in Visual Prompt Tuning!

LCA-on-the-Line: Benchmarking Out of Distribution Generalization with Class Taxonomies