ServiceNow Research

Sai Rajeswar Mudumba

Sai Rajeswar Mudumba

Research Scientist

Low Data Learning

Sai Rajeswar is a Senior Research Scientist at ServiceNow Research, Montreal team. He completed his Ph.D. in Computer Science at Mila in Université de Montréal, under the supervision of Prof. Aaron Courville. During his Ph.D., he spent time at DeepMind as a Research Scientist intern and at ElementAI as a visiting researcher. Earlier, he graduated from IIT Delhi, with masters in Computer Science. He worked as a Research Engineer at Xerox Research Centre before joining for his Ph.D. Previously, he also graduated with an MSc in Mathematics with a focus on Computer Science in India. His research has been focused on perception and control from limited supervision. Topics of interest include representation learning, generative interactive models, and transfer in RL. His research has been published in ICML, ICLR, CVPR, and NeurIPS.

Interests
  • Reinforcement Learning
  • Generative Models
  • Machine Learning

Publications

Mastering the Unsupervised Reinforcement Learning Benchmark from Pixels. Sai Rajeswar Mudumba, Pietro Mazzaglia, Tim Verbelen, Alexandre Piche, Bart Dhoedt, Aaron Courville, Alexandre Lacoste. At International Conference on Machine Learning (ICML),  2023.

Cite

Choreographer: Learning and Adapting Skills in Imagination. Pietro Mazzaglia, Tim Verbelen, Bart Dhoedt, Alexandre Lacoste, Sai Rajeswar Mudumba. At International Conference of Learning Representations (ICLR),  2023.

PDF Cite

Choreographer: Learning and Adapting Skills in Imagination . Pietro Mazzaglia, Tim Verbelen, Bart Dhoedt, Alexandre Lacoste, Sai Rajeswar Mudumba. At Workshop at the Neural Information Processing Systems (NeurIPS),  2022.

PDF Cite

Consistency-CAM: Towards Improved Weakly Supervised Semantic Segmentation. Sai Rajeswar Mudumba, Issam H. Laradji, Pau Rodriguez, David Vazquez, Aaron Courville. At British Machine Vision Conference (BMVC),  2022.

PDF Cite

Unsupervised Model-based Pre-training for Data-efficient Reinforcement Learning from Pixels. Sai Rajeswar Mudumba, Pietro Mazzaglia, Tim Verbelen, Alexandre Piche, Aaron Courville, Alexandre Lacoste. At Workshop at the International Conference on Machine Learning (ICML),  2022.

PDF Cite

Multi-label Iterated Learning for Image Classification with Label Ambiguity. Sai Rajeswar Mudumba, Pau Rodriguez, Soumye Singhal, David Vazquez, Aaron Courville. At Computer Vision and Pattern Recognition (CVPR),  2022.

PDF Cite Code

Adversarial Computation of Optimal Transport Maps. Jacob Leygonie, Jennifer She, Amjad Almahairi, Sai Rajeswar Mudumba, Aaron Courville. At ArXiv,  2019.

PDF Cite Code

Pix2Shape: Towards Unsupervised Learning of 3D Scenes from Images using a View-based Representation. Sai Rajeswar Mudumba, Fahim Mannan, Florian Golemo, Jérôme Parent-Lévesque, David Vazquez, Derek Nowrouzezahrai, Aaron Courville. At International Journal in Computer Vision (IJCV),  2019.

PDF Cite Code

Towards Text Generation with Adversarially Learned Neural Outlines. Sandeep Subramanian, Sai Rajeswar Mudumba, Alessandro Sordoni, Adam Trischler, Aaron Courville, Christopher Pal. At Conference on Neural Information Processing Systems (NeurIPS),  2018.

PDF Cite

Hierarchical Adversarially Learned Inference. Ishmael Belghazi, Sai Rajeswar Mudumba, Olivier Mastropietro, Negar Rostamzadeh, Jovana Mitrovic, Aaron Courville. At Workshop at the International Conference on Machine Learning (ICML),  2018.

PDF Cite Code

Deep Hyperbolic Reinforcement Learning for Continuous Control. Omar Salemohamed, Edoardo Cetin, Sai Rajeswar Mudumba, Arnab Mondal. At ICLR 2023, Tiny Papers,  2017.

Cite