Huiyu Wang

papers

2,245

total citations

papers (19)

Axial-DeepLab: Stand-Alone Axial-Attention for Panoptic Segmentation

ECCV 2020arXiv

789

citations

MaX-DeepLab: End-to-End Panoptic Segmentation With Mask Transformers

CVPR 2021arXiv

597

citations

Ego-Exo4D: Understanding Skilled Human Activity from First- and Third-Person Perspectives

CVPR 2024arXiv

343

citations

Huiyu Wang

papers (19)

Axial-DeepLab: Stand-Alone Axial-Attention for Panoptic Segmentation

MaX-DeepLab: End-to-End Panoptic Segmentation With Mask Transformers

Ego-Exo4D: Understanding Skilled Human Activity from First- and Third-Person Perspectives

CMT-DeepLab: Clustering Mask Transformers for Panoptic Segmentation

Diffusion Models as Masked Autoencoders

Masked Autoencoders Enable Efficient Knowledge Distillers

TubeFormer-DeepLab: Video Mask Transformer

PerceptionLM: Open-Access Data and Models for Detailed Visual Understanding

A Simple Data Mixing Prior for Improving Self-Supervised Learning

CP2: Copy-Paste Contrastive Pretraining for Semantic Segmentation

Ego-Only: Egocentric Action Detection without Exocentric Transferring

k-Means Mask Transformer

SMAUG: Sparse Masked Autoencoder for Efficient Video-Language Pre-Training

MusicFlow: Cascaded Flow Matching for Text Guided Music Generation

BIMBA: Selective-Scan Compression for Long-Range Video Question Answering

In Defense of Image Pre-training for Spatiotemporal Recognition

Learning to Segment Referred Objects from Narrated Egocentric Videos

Ego4D Goal-Step: Toward Hierarchical Understanding of Procedural Activities

HT-Step: Aligning Instructional Articles with How-To Videos

papers (19)

Axial-DeepLab: Stand-Alone Axial-Attention for Panoptic Segmentation

MaX-DeepLab: End-to-End Panoptic Segmentation With Mask Transformers

Ego-Exo4D: Understanding Skilled Human Activity from First- and Third-Person Perspectives

CMT-DeepLab: Clustering Mask Transformers for Panoptic Segmentation

Diffusion Models as Masked Autoencoders

Masked Autoencoders Enable Efficient Knowledge Distillers

TubeFormer-DeepLab: Video Mask Transformer

PerceptionLM: Open-Access Data and Models for Detailed Visual Understanding

A Simple Data Mixing Prior for Improving Self-Supervised Learning

CP2: Copy-Paste Contrastive Pretraining for Semantic Segmentation

Ego-Only: Egocentric Action Detection without Exocentric Transferring

k-Means Mask Transformer

SMAUG: Sparse Masked Autoencoder for Efficient Video-Language Pre-Training

MusicFlow: Cascaded Flow Matching for Text Guided Music Generation

BIMBA: Selective-Scan Compression for Long-Range Video Question Answering

In Defense of Image Pre-training for Spatiotemporal Recognition

Learning to Segment Referred Objects from Narrated Egocentric Videos

Ego4D Goal-Step: Toward Hierarchical Understanding of Procedural Activities

HT-Step: Aligning Instructional Articles with How-To Videos