Bharath Hariharan

I am an associate professor in Computer Science at Cornell University. I work on computer vision and machine learning, in particular on important problems that defy the "Big Data" label. I enjoy problems that require marrying advances in machine learning with insights from computer vision, geometry and domain-specific knowledge. A sampling of the research problems my group works on is presented below; an exhaustive list of publications is available on scholar.


My work has been recognized with an NSF CAREER award and a PAMI Young Researcher Award.

My CV is here and my research statement is here.



Note to prospective PhD students: Admissions at Cornell are done through a committee. If you are interested in working with me, please directly apply through the application website and mention my name
Associate Professor
311 Gates Hall
Cornell University
bharathh-AT-cs-DOT-cornell-DOT-edu

Teaching

PhD students

Former PhD students

Research

Recognition for satellite images and earth science

A variety of scientific disciplines, including environmental science and the earth sciences, need to know what is there in any place on the planet at any time. This requires recognition on satellite images as well as combining information from multiple modalities (satellite, aerial and ground) captured at the same location. Recognition on satellite images is in itself also a fundamental challenge given the absence of large labeled datasets. As part of this project, we have built one of the most accurate foundation vision-language model for satellite images as well as new self-supervised representations for satellite images.

4D Reconstruction and recognition

Humans live in a 4D world: we do not perceive independent static images, but rather a continuous video stream. On the one hand, ego-motion in video provides enough information to reconstruct the static scene and segment the moving objects, which can power recognition. On the other hand, videos depict dynamic scenes and moving objects introduce fundamental ambiguities and challenges with occlusion. Our work has shown how one can resolve ambiguities to reconstruct and segment out moving objects, as well as track surfaces through occlusion.

Representative recent publications


    Remote sensing vision-language foundation models without annotations via ground remote alignment
    Utkarsh Mall, Cheng Perng Phoo, Meilin Kelsey Liu, Carl Vondrick, Bharath Hariharan, Kavita Bala
    In ICLR 2024.
    paper    bibtex
    Dynamo-Depth: Fixing Unsupervised Depth Estimation for Dynamical Scenes
    Yihong Sun, Bharath Hariharan
    In NeurIPS 2023.
    paper    bibtex
    Tracking Everything Everywhere All at Once
    Qianqian Wang, Yen-Yu Chang, Ruojin Cai, Zhengqi Li, Bharath Hariharan, Aleks, er Holynski, Noah Snavely
    In ICCV 2023 (Best Student Paper)
    paper    bibtex
    Emergent Correspondence from Image Diffusion
    Luming Tang, Menglin Jia, Qianqian Wang, Cheng Phoo, Bharath Hariharan
    In NeurIPS 2023.
    paper    bibtex
    Doppelgangers: Learning to Disambiguate Images of Similar Structures
    Ruojin Cai, Joseph Tung, Qianqian Wang, Hadar Averbuch-Elor, Bharath Hariharan, Noah Snavely
    In ICCV 2023 (Oral)
    paper    bibtex
    Change-Aware Sampling and Contrastive Learning for Satellite Images
    Utkarsh Mall, Bharath Hariharan, Kavita Bala
    In CVPR 2023.
    paper    bibtex
    Visual Prompt Tuning
    Menglin Jia, Luming Tang, Bor-Chun Chen, Claire Cardie, Serge Belongie, Bharath Hariharan, Ser-Nam Lim
    In ECCV 2022.
    paper    bibtex
    Learning to Detect Mobile Objects from LiDAR Scans Without Labels
    Yurong You, Katie Luo, Cheng Perng Phoo, Wei-Lun Chao, Wen Sun, Bharath Hariharan, Mark Campbell, Kilian Weinberger
    In CVPR 2022.
    paper    bibtex
    Geometry Processing using Neural Fields
    Guandao Yang, Serge Belongie, Bharath Hariharan, Vladlen Koltun
    In NeurIPS 2021.
    paper    bibtex