I am a PhD student at Max Planck Institute for Intelligent System with Michael Black.
The focus of my research is to develop systems that can learn about the world by seeing them. I work on incorporating geometry about the
world into deep learning systems that solve vision and 3D problems. Some of these problems address Optical Flow, Depth
and Motion Segmentation from both supervised and unsupervised learning perspectives. I also like to work on 3D, especially face meshes. Some
of my work includes incorporating the geometry of the mesh structure within deep learning systems.
I received my Masters degree from the Computer Science Department at The University of British Columbia,
Vancouver. I was a research assistant in the Sensorimotor Systems Lab at UBC
and I worked with Dinesh Pai on understanding motion of Eyes and Upper Faces. I completed my undergraduate
studies at National Institute of Technology Karnataka in the Electronics and Communications Department.
Research Intern, NVIDIA Research
Research Intern, Applied Machine Learning, Facebook Research
PhD Student, Max Planck Institute for Intelligent Systems
Software Developer, Mashup Machine Inc., Vancouver
2013 - 2015
Masters Student, Computer Science, University of British Columbia
Research Intern, Ecole Polytechnique de Montreal
Research Intern, India Innovation Labs, Bangalore
2009 - 2013
Undergrad, National Institute of Technology, Karnataka
Joint Unsupervised Learning of Depth, Camera Motion, Optical Flow and Motion Segmentation
Anurag Ranjan, Varun Jampani, Kihwan Kim, Deqing Sun, Jonas Wulff, Michael J. Black
We address the unsupervised learning of several interconnected problems in low-level vision: single view depth prediction, camera motion estimation, optical flow and segmentation
of a video into the static scene and moving regions. Our model is trained without any supervision and achieves state of the art results amongst unsupervised methods.
Generating 3D Faces using Convolutional Mesh Autoencoders
Anurag Ranjan, Timo Bolkart, Soubhik Sanyal, Michael J. Black (ECCV 2018)
A non-linear model for generating 3D faces using a Convolutional Autoencoder that operates directly on meshes. Our model is state of the art in generating diverse
range of 3D facial meshes.
Unsupervised Learning of Multi-Frame Optical Flow with Occlusions
Joel Janai, Fatma Güney, Anurag Ranjan, Michael J. Black and Andreas Geiger (ECCV 2018)
We propose a framework for unsupervised learning of optical flow and occlusions over multiple frames. Our multi-frame, occlusion-sensitive formulation outperforms existing unsupervised two-frame methods and even produces results on par with some fully supervised methods.
Learning Human Optical Flow
Anurag Ranjan, Javier Romero, Michael J. Black (BMVC 2018)
Learning optical flow for humans is difficult. So, we created a synthetic dataset with realistic humans and trained a neural network on it.
SPyNet: Spatial Pyramid Network for Optical Flow
Anurag Ranjan, Michael J. Black (CVPR 2017)
SPyNet is the smallest deep network in the world that computes optical flow. It is smaller than Flownet by 97% and outperforms it significanly. Both code and trained models are available.
Interactive Gaze Driven Animation of the Eye Region
Debanga R Neog, João L Cardoso, Anurag Ranjan, Dinesh K Pai (Web3D 2016)
A system for real-time animation of eyes that can be interactively controlled in a WebGL. This is the first system for real-time animation of soft tissue movement around the eyes based on gaze input.
Learning Periorbital Soft Tissue Motion
Anurag Ranjan (Master's Thesis, UBC 2015)
We model the soft tissues around the eyes that are associated with subtle and fast motions and convey emotions during facial expressions. Our data driven model that can efficiently learn and reproduce the complex motion of these periorbital soft tissues.
Implementation of 3D object recognition and tracking
This object recognition and tracking system utilizes the depth information from a low-cost depth sensor. This approach makes use of the depth information and 3d properties of objects inorder to accurately identify them independent of lighting conditions.
My research focuses on Deep Learning for dense estimation problems such as Optical Flow. My broad interests include Computer Vision and Machine Learning. My full CV is available here .