I am a research scientist in Meta GenAI. I now mainly work on media foundation models (emu) and multimodal foundation models (llama). I previously worked on depth estimation, on-device computer vision and human vision-inspired computer vision. Prior to Meta, I obtained my Ph.D. from Harvard University advised by Prof. Todd Zickler and my B.A.Sc from the University of Toronto advised by Prof. Sven Dickinson and Prof. Sanja Fidler.
📩 Email: jialiang.wang at alumni.harvard.edu
💻 Selected Publications
GenAI
Cache Me if You Can: Accelerating Diffusion Models through Block Caching
F. Wimbauer, B. Wu, E. Schoenfeld, X. Dai, J. Hou, Z. He, A. Sanakoyeu, P. Zhang, S. Tsai, J. Kohler, C. Rupprecht, D. Cremers, P. Vajda, J. Wang
CVPR, Seattle, WA, June 2024Emu: Enhancing Image Generation Models Using Photogenic Needles in a Haystack
X. Dai∗, J. Hou∗, CY Ma∗, S Tsai∗, J. Wang∗, R. Wang∗, P. Zhang∗, S. Vandenhende, X. Wang, A. Dubey, M. Yu, A. Kadian, F. Radenovic, D. Mahajan, K. Li, Y. Zhao, V. Petrovic, M. K. Singh, S. Motwani, Y. Wen, Y. Song, R. Sumbaly†, V. Ramanathan†, Z. He†, P. Vajda†, D. Parikh†
Meta AI Tech Report, 2023
∗: Equal contribution: alphabetical order
†: joint last authors
Depth estimation and human depth perception
A Practical Stereo Depth System for Smart Glasses
J. Wang, D. Scharstein, A. Bapat, K. Blackburn-Matzen, M. Yu, J. Lehman, S. Alsisan, Y. Wang, S. Tsai, JM Frahm, Z. He, P. Vajda, M. F. Cohen, M. Uyttendaele
CVPR, Vancouver, BC, June 2023Toward practical monocular indoor depth estimation
CY Wu, J. Wang, M. Hall, U. Neumann, S. Su
CVPR, New Orleans, LA, June 2022A lighting-invariant point processing for shading
K. Heal, J. Wang, S. J. Gortler, T. Zickler
CVPR, Seattle, WA, June 2020Local detection of stereo occlusion boundaries
J. Wang, T. Zickler
CVPR, Long Beach, CA, June 2019Half-occlusion boundary detectors in computational stereo vision
J. Wang, D. Glasner, T. Zickler,
Journal of Vision, VSS Abstract, St. Pete Beach, FL, May 2018 [slides][ICCV'17 paper][stereoscope viewer]
💻 All Papers
Cache Me if You Can: Accelerating Diffusion Models through Block Caching
F. Wimbauer, B. Wu, E. Schoenfeld, X. Dai, J. Hou, Z. He, A. Sanakoyeu, P. Zhang, S. Tsai, J. Kohler, C. Rupprecht, D. Cremers, P. Vajda, J. Wang
CVPR, 2024FlowVid: Taming Imperfect Optical Flows for Consistent Video-to-Video Synthesis
F. Liang, B. Wu, J. Wang, L. Yu, K. Li, Y. Zhao, I. Misra, JB Huang, P. Zhang, P. Vajda, D. Marculescu
CVPR, 2024ControlRoom3D: Room Generation using Semantic Proxy Rooms
J. Schult, S. Tsai, L. Höllein, B. Wu, J. Wang, CY Ma, K. Li, X. Wang, F. Wimbauer, Z. He, P. Zhang, B. Leibe, P. Vajda, J. Hou
CVPR, 2024Efficient Quantization Strategies for Latent Diffusion Models
Y. Yang, X. Dai, J. Wang, P. Zhang, H. Zhang
CVPR workshop on Efficient and On-Device Generation, 2024An Analysis on Quantizing Diffusion Transformers
Y. Yang, J. Wang, X. Dai, P. Zhang, H. Zhang
arXiv, 2024Emu: Enhancing Image Generation Models Using Photogenic Needles in a Haystack
X. Dai∗, J. Hou∗, CY Ma∗, S Tsai∗, J. Wang∗, R. Wang∗, P. Zhang∗, S. Vandenhende, X. Wang, A. Dubey, M. Yu, A. Kadian, F. Radenovic, D. Mahajan, K. Li, Y. Zhao, V. Petrovic, M. K. Singh, S. Motwani, Y. Wen, Y. Song, R. Sumbaly†, V. Ramanathan†, Z. He†, P. Vajda†, D. Parikh†
Meta AI Tech Report, 2023
∗: Equal contribution: alphabetical order
†: joint last authorsNeRF-Det: Learning Geometry-Aware Volumetric Representation for Multi-View Indoor 3D Object Detection
C. Xu, B. Wu, J. Hou, S. Tsai, R. Li, J. Wang, W. Zhan, Z. He, P. Vajda, K. Keutzer, M. Tomizuka
ICCV, 2023A Practical Stereo Depth System for Smart Glasses
J. Wang, D. Scharstein, A. Bapat, K. Blackburn-Matzen, M. Yu, J. Lehman, S. Alsisan, Y. Wang, S. Tsai, JM Frahm, Z. He, P. Vajda, M. F. Cohen, M. Uyttendaele
CVPR, 2023Consistent Direct Time-of-Flight Video Depth Super-Resolution
Z. Sun, W. Ye, J. Xiong, G. Choe, J. Wang, S. Su, R. Ranjan
CVPR, 2023Toward practical monocular indoor depth estimation
CY Wu, J. Wang, M. Hall, U. Neumann, S. Su
CVPR, 2022FBNetV5: Neural architecture search for multiple tasks in one run
B. Wu, C. Li, H. Zhang, X. Dai, P. Zhang, M. Yu, J. Wang, Y. Lin and P. Vajda
arXiv, 2021Level set binocular stereo with occlusions
J. Wang, T. Zickler
arXiv, 2021Level set stereo for cooperative grouping with occlusion
J. Wang, T. Zickler
ICIP, 2021A lighting-invariant point processing for shading
K. Heal, J. Wang, S. J. Gortler, T. Zickler
CVPR, 2020Interpreting robust optimization via adversarial influence functions
Z. Deng, C. Dwork, J. Wang, L. Zhang
alphabetical order
ICML, 2020Improving deep stereo network generalization with geometric priors
J. Wang, V. Jampani, D. Sun, C. Loop, S. Birchfield, J. Kautz
arXiv, 2020Local detection of stereo occlusion boundaries
J. Wang, T. Zickler
CVPR, 2019A computational model for local stereo occlusion boundary detection
J. Wang, T. Zickler,
Journal of Vision, VSS Abstract, 2019 [poster][project][stereoscope viewer]Half-occlusion boundary detectors in computational stereo vision
J. Wang, D. Glasner, T. Zickler,
Journal of Vision. VSS Abstract, 2018 [slides][project][stereoscope viewer]Toward perceptually-consistent stereo: A scanline study
J. Wang, D.Glasner, T. Zickler
ICCV, 2017
📃 Patents
Wang, J., et al. "Distance determinations using one or more neural networks."
U.S. Patent Application No. 16/852,944
⌨️ Service
Reviewer: CVPR'20-24, NeurIPS'20-24, ICML'21-23, ICCV 21,23, ICLR'21-22, BMVC'20, ACCV'20, WACV'21-22, ECCV'22, 24