Lorenzo Torresani

Associate Professor of Computer Science

Visual Learning Group
Dartmouth College

6211 Sudikoff Lab, Hanover, NH 03755
Fax: (603) 646.


Since summer 2017 I have been splitting my time between Dartmouth and Facebook, where I do research in the area of video understanding.

recent news

  1. BulletI am looking for two postdocs to join me at Facebook AI Research (FAIR) in Boston to work in the area of video understanding, starting in or after Fall 2018. Please contact me to apply.

  2. BulletThree papers on video models to appear at CVPR 2018:

  3.   A Closer Look at Spatiotemporal Convolutions for Action Recognition,
        with Du Tran, Heng Wang, Jamie Ray, Yann LeCun, and Manohar Paluri.

  4.   Detect-and-Track: Efficient Pose Estimation in Videos,
        with Rohit Girdhar, Georgia Gkioxari, Manohar Paluri, and Du Tran.

  5.   What Makes a Video a Video: Analyzing Temporal Information in Video
    Understanding Models and Datasets,
        with De-An Huang, Vignesh Ramanathan, Dhruv Mahajan, Juan Carlos Niebles,
        Fei-Fei  Li, and Manohar Paluri.

  1. BulletTogether with collaborators, I am organizing two workshops at CVPR 2018:

  2. Brave New Ideas for Video Understanding

  3. DeepGlobe: A Challenge for Parsing the Earth through Satellite Images

   We invite paper submissions to both these workshops!

  1. BulletFrom January to May 2018 I will be at Ashesi University in Ghana on a Fulbright U.S. Scholar Award.

  1. BulletTwo new papers on arXiv:

  2.   Connectivity Learning in Multi-Branch Networks,
        with Karim Ahmed.

  3.   VideoMCC: a New Benchmark for Video Comprehension,
        with Du Tran, Maksim Bolonkin, and Manohar Paluri.

  4. Bullet New paper to appear at WACV 2018:

   BranchConnect: Large-Scale Visual Recognition with Learned Branch Connections,
   with Karim Ahmed.

  1. Bullet New paper at NIPS 2017:

   Learning to Inpaint for Image Compression,

   with Haris Baig, and Vladlen Koltun.

  1. Bullet New paper at CVPR 2017:

   Convolutional Random Walk Networks for Semantic Image Segmentation,
   with Gedas Bertasius, Stella Yu and Jianbo Shi.

  1. Bullet New paper at ICLR 2017:

   Recurrent Mixture Density Network for Spatiotemporal Visual Attention,
   with Loris Bazzani and Hugo Larochelle.

  1. Bullet New paper at AISTATS 2017:

   Local Perturb-and-MAP for Structured Prediction,
   with Gedas Bertasius, Qiang Liu and Jianbo Shi.

  1. Bullet New paper to appear in Computer Vision and Image Understanding:

   Multiple Hypothesis Colorization and Its Application to Image Compression,
   with Haris Baig.

  1. Bullet We released our new dataset for video comprehension, named VideoMCC.
       It includes over 600 hours of video. Give it a try!

  1. Bullet Area chair for ICCV 2017.

  1. Bullet Area chair for CVPR 2017.

  1. Bullet Together with collaborators at Facebook and Google I co-organized the
    1st Workshop on Large Scale Computer Vision Systems (LSCVS), at NIPS 2016.

  1. Bullet New paper at SPIE Symposium on Electronic Imaging 2017:

   Deciphering Severely Degraded License Plates,
   with Shruti Agarwal, Du Tran, and Hany Farid.

  1. Bullet New paper at ECCV 2016:

   Network of Experts for Large-Scale Image Categorization,
   with Karim Ahmed and Haris Baig.

  1. Bullet New paper at CVPR 2016:

   Semantic Segmentation with Boundary Neural Fields,
   with Gedas Bertasius and Jianbo Shi.

  1. Bullet New paper at the 3rd Workshop on Deep Learning in Computer Vision, 2016:

   Deep End2End Voxel2Voxel Prediction,
   with Du Tran, Lubomir Bourdev, Rob Fergus and Manohar Paluri.

  1. Bullet New article in IJCV:

   EXMOVES: Mid-level Features for Efficient Action Recognition and Video Analysis,
   with Du Tran.


  1. Bullet Thanks to Nvidia Corporation for supporting our research with a hardware donation.

research overview

My research interests are in computer vision and machine learning. My current work is primarily focused on learning representations for image and video recognition. You can read more about the research of my group here.

previous affiliations

Microsoft Research Cambridge, Machine Learning and Perception
New York University, Computer Science
Stanford University, Computer Science
University of Milan, Computer Science