Shurjo Banerjee
5th year Phd @ The University of Michigan, Ann Arbor
B.S. in Computer Engineering @ The Georgia Institute of Technology, 2015

I am a 5th year PhD at the University of Michigan. I am broadly interested in deep learning and its applications to deep reinforcement learning, incremental learning and vision-and-language (VLN) problems.

News

Accepted Paper
  CORL
The RobotSlang Benchmark: Dialog-guided Robot Localization and NavigationOctober 2020
    

Papers

2020
The RobotSlang Benchmark: Dialog-guided Robot Localization and Navigation
Shurjo Banerjee, Jesse Thomason, and Jason J. Corso.
The Conference on Robot Learning (CORL), 2020.
papervideo
@inproceedings{robotslang:corl20,
  title={The RobotSlang Benchmark: Dialog-guided Robot Localization and Navigation},
  author={Shurjo Banerjee and Jesse Thomason and Jason J. Corso},
  booktitle={The Conference on Robot Learning (CORL)},
  year={2020},
  url={https://arxiv.org/abs/2010.12639}
}
2019
A Geometric Approach to Online Streaming Feature Selection
Salimeh Y. Sekeh, Madan R. Ganesh, Shurjo Banerjee, Jason J. Corso, and Alfred H. Hero.
ArXiv, 2019.
paper
@inproceedings{goa19,
  title={A Geometric Approach to Online Streaming Feature Selection},
  author={Salimeh Y. Sekeh and Madan R. Ganesh and Shurjo Banerjee and Jason J. Corso and Alfred H. Hero},
  booktitle={ArXiv},
  year={2019},
  url={https://arxiv.org/abs/1910.01182}
}
Learning goal-conditioned value functions with one-step path rewards rather than goal rewards
Vikas Dhiman, Shurjo Banerjee, Jeffrey M. Siskind, and Jason J. Corso.
OpenReview, 2019.
paper
@inproceedings{onestepreward18,
  title={Learning goal-conditioned value functions with one-step path rewards rather than goal rewards},
  author={Vikas Dhiman and Shurjo Banerjee and Jeffrey M. Siskind and Jason J. Corso},
  booktitle={OpenReview},
  year={2019},
  url={https://openreview.net/pdf?id=BkesGnCcFX}
}
Floyd-Warshall Reinforcement Learning: Learning from Past Experiences to Reach New Goals
Vikas Dhiman, Shurjo Banerjee, Jeffrey M. Siskind, and Jason J. Corso.
arXiv, 2019.
paper
@inproceedings{floydwarshall18,
  title={Floyd-Warshall Reinforcement Learning: Learning from Past Experiences to Reach New Goals},
  author={Vikas Dhiman and Shurjo Banerjee and Jeffrey M. Siskind and Jason J. Corso},
  booktitle={arXiv},
  year={2019},
  url={https://openreview.net/pdf?id=BkesGnCcFX}
}
2018
A Critical Investigation of Deep Reinforcement Learning for Navigation
Shurjo Banerjee, Vikas Dhiman, Brent Griffin, Jeffrey M. Siskind, and Jason J. Corso.
ArXiv, 2018.
paper
@inproceedings{criticalinvestigation18,
  title={A Critical Investigation of Deep Reinforcement Learning for Navigation},
  author={Shurjo Banerjee and Vikas Dhiman and Brent Griffin and Jeffrey M. Siskind and Jason J. Corso},
  booktitle={ArXiv},
  year={2018},
  url={https://arxiv.org/abs/1802.02274}
}
2017
Do Deep Reinforcement Learning Algorithms really learn to Navigate?
Vikas Dhiman, Shurjo Banerjee, Brent Griffin, and Jason J. Corso.
NeurIps Deep Reinfocement Learning Workshop, 2017.
paper
@inproceedings{dodeepnavigate17,
  title={Do Deep Reinforcement Learning Algorithms really learn to Navigate?},
  author={Vikas Dhiman and Shurjo Banerjee and Brent Griffin and Jason J. Corso},
  booktitle={NeurIps Deep Reinfocement Learning Workshop},
  year={2017},
  url={https://drive.google.com/file/d/1WfC6iJEFyJtw9VL6ppvNz03_kcpx1DKj/view}
}
2016
Towards a wireless multimodal speech capture system
Nordine Sebkhi, Dhyey Desai, Aamir Khan, Nischal Prasad, Shurjo Banerjee, Justin Eng, Kimberly Wilson, and Maysam M. Ghovanloo.
IEEE Biomedical Circuits and Systems Conference, 2016.
paper
@inproceedings{multimodalTTS16,
  title={Towards a wireless multimodal speech capture system},
  author={Nordine Sebkhi and Dhyey Desai and Aamir Khan and Nischal Prasad and Shurjo Banerjee and Justin Eng and Kimberly Wilson and Maysam M. Ghovanloo},
  booktitle={IEEE Biomedical Circuits and Systems Conference},
  year={2016},
  url={https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=\&arnumber=7833730\&tag=1}
}