PPaperPicks

Haofeng Huang

8 papers at tracked venues · 7 at CORE A* · active 20242025

Venues

Frequent coauthors

Papers

  1. Faster Video Diffusion with Trainable Sparse Attention
  2. SageAttention2: Efficient Attention with Thorough Outlier Smoothing and Per-thread INT4 Quantization
  3. SageAttention3: Microscaling FP4 Attention for Inference and An Exploration of 8-Bit Training
  4. SpargeAttention: Accurate and Training-free Sparse Attention Accelerating Any Model Inference
  5. ViDiT-Q: Efficient and Accurate Quantization of Diffusion Transformers for Image and Video Generation
  6. XAttention: Block Sparse Attention with Antidiagonal Scoring
  7. FreeMotion: MoCap-Free Human Motion Synthesis with Multimodal Large Language Models
  8. Seeing Dark Videos via Self-Learned Bottleneck Neural Representation
    AAAI 2024 · Haofeng Huang