Cvpr 2026 Unipr

UniPR

In this video, we introduce a novel video object detection framework called D2FANet. D2FANet is the first framework to jointly ...

Rameen Abdal, James Burgess, Sergey Tulyakov, Kuan-Chieh Wang Snap Research , Stanford University ...

Disentangle-then-Align: Non-Iterative Hybrid Multimodal Image Registration via Cross-Scale Feature Disentanglement.

Poster Presentation

Description: This is the official demo video for the paper "UniSER: A Foundation Model for Unified Soft Effects Removal", accepted ...

https://arxiv.org/abs/2511.20648.

TAPE: Task-Adaptive Prototype Evolution in Audio-Language Models for Fully Few-shot Class-incremental Audio Classification.

VIMCAN: Visual-Inertial 3D Human Pose Estimation with Hybrid Mamba-Cross-Attention Network.

CVPR 2026

Despite significant progress has been made in image deraining, we note that most existing methods are often developed for only ...

CVPR 2026

Omni-Attribute encodes a high-fidelity, attribute-specific image representation, that enables coherent synthesis of the ...

Paper: https://arxiv.org/abs/2512.09270 Project Page: https://cmlab-korea.github.io/MoRel Authors/Affiliations: [Sangwoon ...

Leon Liangyu Chen, Haoyu Ma, Zhipeng Fan, Ziqi Huang, Animesh Sinha, Xiaoliang Dai, Jialiang Wang, Zecheng He, Jianwei ...

[

Paper: https://arxiv.org/abs/2512.01390 Project Page: https://cmlab-korea.github.io/FRAMER/ Authors/Affiliations: [Seungho ...

A year with 100+ content creators teaching AI to describe video like a filmmaker — with a structured specification, scalable ...