Vincent Casser

Machine Learning Researcher, Software Engineer


I am a current graduate student in Computational Science at Harvard University, with major interests in machine learning and visual computing. My previous research was mostly related to computer vision: in domains such as autonomous UAV navigation, autonomous driving, traffic scene segmentation and biomedical imaging. I have high interests in interdisciplinary applications, with my previous projects including collaborations with representatives of Massachusetts General Hospital in Healthcare, the Harvard-Smithsonian Center in Astrophysics, and the Harvard Lichtman Lab in Neuroscience.

I’ve been an avid programmer for 10 years now and have a comprehensive background in the Microsoft.NET domain. I was one of the main contributors to the largest German-speaking .NET online community for several years. I also have experience with Python, C/C++, C# and Java, and I taught several courses as a teaching fellow (024 OOSE ST15014 ADIP WT15024 OOSE ST16). Before doing computer science research, I developed various software applications for non-scientific purposes.

I graduated with a BSc in Computer Science from the University of Bonn, Germany in 2016, where I minored in physics and astronomy. I then worked as a research intern at King Abdullah University of Science and Technology (KAUST), and joined the Harvard John A. Paulson School of Engineering and Applied Sciences (SEAS) in Fall 2017 as a Master’s student, from which I will graduate in May 2019. During the summer of 2018, I interned in the Google Brain Robotics group under Anelia Angelova. I recently joined Aude Oliva’s lab at MIT CSAIL, where I will work on (multimodal) video understanding. I use this webspace for hosting personal projects, and plan to publish some of them publicly here.


02/01/19: Presented our struct2depth work at AAAI-19 in Hawaii
01/23/19: Gave the workshop on “Convolutional Autoencoders for Image Manipulation” at ComputeFest 2019
12/16/18: Preprint of our work on connectomics now on arXiv
11/28/18: New project OIL: Observational Imitation Learning now released
11/27/18: New blog post on our struct2depth work on Google’s AI blog
11/19/18: The code for our struct2depth paper is now part of the TensorFlow models repository: Link
11/01/18: One paper accepted to AAAI’19: “Depth Prediction Without the Sensors: Leveraging Structure for Unsupervised Learning from Monocular Videos” (full text)
10/06/18: Joined the MIT Computational Perception & Cognition Lab, led by Aude Olivia
09/08/18: Our paper won the best paper award at UAVISION 2018
08/03/18: We are presenting our work on autonomous drone racing on Sept 8 at UAVISION, ECCV’18
05/29/18: Started internship in the Google Brain Robotics group
05/22/18: Our new datasets for connectomics research are now publicly available: Kasthuri++ and Lucchi++
05/21/18: Release of new tutorial for Bayesian GAN
02/08/18: Started new project on Connectomics with the Visual Computing Group (VCG)
01/22/18: Started new collaboration with the Center for Clinical Data Science (CCDS)
11/23/17: One paper accepted to IJCV: “Sim4CV: A Photo-Realistic Simulator for Computer Vision Applications” (full text)
10/21/17: Official release of Sim4CV, our simulation environment for Computer Vision
09/01/17: Started Master’s program in Computational Science and Engineering
08/19/17: New paper released on arXiv: “Autonomous UAV Racing Using Deep Learning
05/24/17: Recipient of German Academic Scholarship Foundation US-Scholarship (Studienstiftung)
03/28/17: Recipient of DAAD Graduate Scholarship