Progressive Mixed-Precision Decoding for Efficient LLM Inference
Hao Mark Chen, Fuwen Tan, Alexandros Kouris, Royson Lee, Hongxiang Fan, Stylianos I. Venieris
Fuwen Tan
I work on making image/video generation models and large language models faster, with recent work on distillation, quantization, and efficient inference. I am a main developer of MAI-Image-2.5-Flash, an efficiency-optimized variant of MAI-Image-2.5, and received a Best Paper Finalist at CVPR 2019.
Research
MobileQuant: Mobile-friendly Quantization for On-device Language Models
Fuwen Tan, Royson Lee, Lukasz Dudziak, Shell Xu Hu, Sourav Bhattacharya, Timothy Hospedales, Georgios Tzimiropoulos, Brais Martinez
iBoot: Image-bootstrapped Self-Supervised Video Representation Learning
Fatemeh Saleh, Fuwen Tan, Adrian Bulat, Georgios Tzimiropoulos, Brais Martinez
EdgeViTs: Competing Light-weight CNNs on Mobile Devices with Vision Transformers
Junting Pan, Adrian Bulat, Fuwen Tan, Xiatian Zhu, Lukasz Dudziak, Hongsheng Li, Georgios Tzimiropoulos, Brais Martinez
Drill-down: Interactive Retrieval of Complex Scenes using Natural Language Queries
Fuwen Tan, Paola Cascante-Bonilla, Xiaoxiao Guo, Hui Wu, Song Feng, Vicente Ordonez
Text2Scene: Generating Compositional Scenes from Textual Descriptions
Fuwen Tan, Song Feng, Vicente Ordonez
Best Paper Finalist
Where and Who? Automatic Semantic-Aware Person Composition
Fuwen Tan, Crispin Bernier, Benjamin Cohen, Vicente Ordonez, Connelly Barnes
FaceCollage: A Rapidly Deployable System for Real-time Head Reconstruction for On-The-Go 3D Telepresence
Fuwen Tan, Chi-Wing Fu, Teng Deng, Jianfei Cai, Tat Jen Cham
High-Quality Kinect Depth Filtering For Real-time 3D Telepresence
Mengyao Zhao, Fuwen Tan, Chi-Wing Fu, Chi-Keung Tang, Jianfei Cai, Tat Jen Cham
Thesis
Learning Local Representations of Images and Text
Images and text exhibit hierarchical structures: scenes are built from objects, sentences from words. This thesis develops techniques for learning local representations of images and text, with applications in visual recognition, retrieval, and synthesis.