Kinal Mehta

Bangalore

I am a research engineer at Qualcomm XR, focusing on Human body animation utilizing Reinforcement Learning (RL) techniques.

My academic journey includes earning an MS by research from IIIT Hyderabad, where I developed “marl-jax,” a MARL (Multi-Agent Reinforcement Learning) framework with a specific emphasis on achieving social generalization.

Beyond my professional commitments, I actively contribute to the open-source community. You can find my contributions to projects such as Cleanrl and other RL libraries during my free time.

Prior to my time at IIIT Hyderabad, I gained valuable experience working on diverse projects in the realms of Natural Language Processing (NLP) and Computer Vision.

I’m passionate about pushing the boundaries of technology and applying learning algorithms to solve real-world challenges.

selected publications

ECML-PKDD

marl-jax: Multi-agent Reinforcement Leaning Framework for Social Generalization

Mehta, Kinal, Mahajan, Anuj, and Kumar, Pawan

In European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases. Demo Track 2023

Bib HTML

@inproceedings{mehta2023marl-jax,
  abbr = {ECML-PKDD},
  author = {Mehta, Kinal and Mahajan, Anuj and Kumar, Pawan},
  title = {marl-jax: Multi-agent Reinforcement Leaning Framework for Social Generalization},
  booktitle = {European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases. Demo Track},
  year = {2023},
  publisher = {Springer International Publishing},
  html = {https://arxiv.org/abs/2303.13808},
  bibtex_show = {true},
  selected = {true}
}

IEEE IJCNN

Effects of Spectral Normalization in Multi-agent Reinforcement Learning

Mehta, Kinal, Mahajan, Anuj, and Kumar, Pawan

In International Joint Conference on Neural Networks (IJCNN) 2023

Bib HTML

@inproceedings{mehta2023effects,
  abbr = {IEEE IJCNN},
  author = {Mehta, Kinal and Mahajan, Anuj and Kumar, Pawan},
  booktitle = {International Joint Conference on Neural Networks (IJCNN)},
  title = {Effects of Spectral Normalization in Multi-agent Reinforcement Learning},
  year = {2023},
  html = {https://arxiv.org/abs/2212.05331},
  bibtex_show = {true},
  selected = {true}
}

JMLR

CleanRL: High-quality Single-file Implementations of Deep Reinforcement Learning Algorithms

Huang, Shengyi, Dossa, Rousslan Fernand Julien, Ye, Chang, Braga, Jeff, Chakraborty, Dipam, Mehta, Kinal, and AraÃºjo, JoÃ£o G.M.

Journal of Machine Learning Research 2022

Bib HTML

@article{JMLR:v23:21-1342,
  abbr = {JMLR},
  author = {Huang, Shengyi and Dossa, Rousslan Fernand Julien and Ye, Chang and Braga, Jeff and Chakraborty, Dipam and Mehta, Kinal and AraÃºjo, JoÃ£o G.M.},
  title = {CleanRL: High-quality Single-file Implementations of Deep Reinforcement Learning Algorithms},
  journal = {Journal of Machine Learning Research},
  booktitle = {Journal of Machine Learning Research},
  year = {2022},
  volume = {23},
  number = {274},
  pages = {1--18},
  html = {http://jmlr.org/papers/v23/21-1342.html},
  bibtex_show = {true},
  selected = {true}
}