Tao Huang
Assistant Professor
Works on efficient deep learning and computer vision, including model compression and knowledge distillation.
Tao Huang's lab at SJTU works on efficient training & inference and knowledge, QA & retrieval. 28 papers since 2023, 24 at top venues, mostly ECCV and CVPR. 3 PhD students and 3 master's students. Applications close .
At a glance
- papers, 2023–2026
- 7/ year
- at T1 venues
- 86%
- PhD students
- 3
- MS students
- 3
- alumni listed
- —
Papers by yearfrom the lab homepage, 2023–2026
- 13 papers2026 · T1 12 · collecting
- 6 papers2025 · T1 6
- 6 papers2024 · T1 4
- 3 papers2023 · T1 2
T1 venuesOtherVenues: ECCV, CVPR, ICLR, NeurIPS, AAAI
Research focuspaper titles that use each topic's words
- Knowledge, QA & retrieval8 papers
- Efficient training & inference7 papers
- distillation7 papers
- model compression0 papers
- efficient deep0 papers
Paper list
From the lab homepage.
202613 papers
- CVPR (T1)Affordance Field Intervention: Enabling VLAs to Escape Memory Traps in Robotic Manipulation (opens in a new tab)
- ECCV (T1)AutoV: Loss-Oriented Ranking for Visual Prompt Retrieval in LVLMs (opens in a new tab)
- ECCV (T1)DASH: Dynamic Audio-Driven Semantic Chunking for Efficient Omnimodal Token Compression (opens in a new tab)
- AAAI (T1)Distilling Cross-Modal Knowledge via Feature Disentanglement (opens in a new tab)
- CVPR (T1)Do you have freestyle? expressive humanoid locomotion via audio control (opens in a new tab)
- TPAMI (T1)FreeKD+: A Frequency Knowledge Distillation Framework for Dense Prediction (opens in a new tab)
- ICLR (T1)From Language to Locomotion: Retargeting-free Humanoid Control via Motion Latent Guidance (opens in a new tab)
- ACM MMLearning to Expand Images for Efficient Visual Autoregressive Modeling (opens in a new tab)
- ECCV (T1)Pistachio: Towards Synthetic, Balanced, and Long-Form Video Anomaly Benchmarks (opens in a new tab)
- ICLR (T1)Rethinking Causal Mask Attention for Vision-Language Inference (opens in a new tab)
- ECCV (T1)Revisiting Parameter Redundancy in Vision-Language-Action Models: Insights from VLM-to-VLA Adaptation (opens in a new tab)
- ECCV (T1)RoboMirror: Understand Before You Imitate for Video to Humanoid Locomotion (opens in a new tab)
20256 papers
- TPAMI (T1)DIST+: Knowledge Distillation from A Stronger Adaptive Teacher (opens in a new tab)
- AAAI (T1)EfficientVMamba: Atrous Selective Scan for Light Weight Visual Mamba (opens in a new tab)
- ICLR (T1)Learning Mask Invariant Mutual Information for Masked Image Modeling (opens in a new tab)
- CVPR (T1)MoVE-KD: Knowledge Distillation for VLMs with Mixture of Visual Encoders (opens in a new tab)
- ICML (T1)SparseVLM: Visual Token Sparsification for Efficient Vision-Language Model Inference (opens in a new tab)
- NeurIPS (T1)VLA-Cache: Efficient Vision-Language-Action Manipulation via Adaptive Token Caching (opens in a new tab)
20246 papers
- ECCV (T1)Active Generation for Image Classification (opens in a new tab)
- Knowledge-Based SystemsDiverse and Tailored Image Generation for Zero-Shot Multi-Label Classification (opens in a new tab)
- CVPR (T1)FreeKD: Knowledge Distillation via Semantic Frequency Prompt (opens in a new tab)
- ICMEPositive Label Is All You Need for Multi-Label Classification (opens in a new tab)
- ECCV (T1)Training-free Composite Scene Generation for Layout-to-Image Synthesis (opens in a new tab)
- NeurIPS (T1)Unveiling the Tapestry of Consistency in Large Vision-Language Models (opens in a new tab)
Where alumni went
Alumni are not listed on the lab homepage.
Apply to SJTUFall 2027
Computer Science PhD (international applicants)
- Round
- Self-funded
- Opens
- —
- GRE
- See program page
- English
- See program page
- Fee
- See program page
Not yet confirmed for this intake.
Similar labsnearest by research summary
- CHETH Zurich
Florian Tramèr
- Trustworthy & safe ML
- SGNUS
Mehul Motani
- Signal processing & information theory
- GBOxford
Bernardo Cuenca Grau
- Reasoning & neuro-symbolic AI
- GBUCL
John Shawe-Taylor
- Learning theory & optimization
- USUCSD
Tianhao Wang
- Learning theory & optimization
- USUIUC
Venugopal V. Veeravalli
- Learning theory & optimization
Your notes
Save this lab to keep notes.
Data
Some numbers could not be verified:
- The alumni list could not be found on the lab site.
- Collected
- Verified
- Papers
- The lab homepage
- Members
- Lab homepage
- Alumni
- Not collected
Tiers are venue classifications, not lab rankings.
SJTUcloses