AI & ML interests
Efficient algorithm, system, and hardware for machine learning.
Recent Activity
View all activity
VILA-U: a Unified Foundation Model Integrating Visual Understanding and Generation
Efficient Diffusion Models with Deep Compression Autoencoder
-
Deep Compression Autoencoder for Efficient High-Resolution Diffusion Models
Paper • 2410.10733 • Published • 9 -
mit-han-lab/dc-ae-f32c32-in-1.0-dit-xl-in-512px
Updated • 6 • 9 -
mit-han-lab/dc-ae-f32c32-in-1.0-uvit-h-in-512px
Updated • 1 • 2 -
mit-han-lab/dc-ae-f32c32-in-1.0-uvit-s-in-512px
Updated • 1
Locality-aware Parallel Decoding for Efficient Autoregressive Image Generation
Models and datasets for "SVDQuant: Absorbing Outliers by Low-Rank Components for 4-Bit Diffusion Models"
Deep Compression Autoencoder
-
EfficientViT: Lightweight Multi-Scale Attention for On-Device Semantic Segmentation
Paper • 2205.14756 • Published • 1 -
EfficientViT-SAM: Accelerated Segment Anything Model Without Performance Loss
Paper • 2402.05008 • Published • 23 -
mit-han-lab/efficientvit-sam
Mask Generation • Updated • 41
Locality-aware Parallel Decoding for Efficient Autoregressive Image Generation
Models and datasets for "SVDQuant: Absorbing Outliers by Low-Rank Components for 4-Bit Diffusion Models"
VILA-U: a Unified Foundation Model Integrating Visual Understanding and Generation
Deep Compression Autoencoder
Efficient Diffusion Models with Deep Compression Autoencoder
-
Deep Compression Autoencoder for Efficient High-Resolution Diffusion Models
Paper • 2410.10733 • Published • 9 -
mit-han-lab/dc-ae-f32c32-in-1.0-dit-xl-in-512px
Updated • 6 • 9 -
mit-han-lab/dc-ae-f32c32-in-1.0-uvit-h-in-512px
Updated • 1 • 2 -
mit-han-lab/dc-ae-f32c32-in-1.0-uvit-s-in-512px
Updated • 1
-
EfficientViT: Lightweight Multi-Scale Attention for On-Device Semantic Segmentation
Paper • 2205.14756 • Published • 1 -
EfficientViT-SAM: Accelerated Segment Anything Model Without Performance Loss
Paper • 2402.05008 • Published • 23 -
mit-han-lab/efficientvit-sam
Mask Generation • Updated • 41