Nikhil Mishra

nikhilmishra000@gmail.com

I am a research scientist at Amazon, where I work on foundation models for robotics that span vision, language, and manipulation. Before that, I was a founding research scientist at Covariant, a startup developing AI for robotics that was reverse-acquihired by Amazon. While working at Covariant, I did my PhD at UC Berkeley, where I was advised by Pieter Abbeel. Before that, I also spent time at OpenAI and Google Brain.

cv   google scholar   linkedin

Publications:

Closing the Visual Sim-to-Real Gap with Object-Composable NeRFs (arxiv)
Nikhil Mishra, Maximilian Sieb, Pieter Abbeel, Xi Chen.
In the proceedings of the International Conference on Robotics and Automation (ICRA), 2024.

Convolutional Occupancy Models for Dense Packing of Complex, Novel Objects (arxiv)
Nikhil Mishra, Pieter Abbeel, Xi Chen, Maximilian Sieb.
In the proceedings of the International Conference on Intelligent Robots and Systems (IROS), 2023.

Distributional Instance Segmentation: Modeling Uncertainty and High Confidence Predictions with Latent-MaskRCNN (arxiv)
YuXuan Liu, Nikhil Mishra, Pieter Abbeel, Xi Chen.
In the proceedings of the International Conference on Robotics and Automation (ICRA), 2023.

Autoregressive Uncertainty Modeling for 3D Bounding Box Prediction (arxiv)
YuXuan Liu, Nikhil Mishra, Maximilian Sieb, Yide Shentu, Pieter Abbeel, Xi Chen.
In the proceedings of the European Conference on Computer Vision (ECCV), 2022.

PixelSNAIL: An Improved Autoregressive Generative Model (arxiv)
Xi Chen, Nikhil Mishra, Mostafa Rohaninejad, Pieter Abbeel.
In the proceedings of the International Conference on Machine Learning (ICML), 2018.
A preliminary version appeared as a workshop paper at ICLR 2018.

A Simple Neural Attentive Meta-Learner (arxiv)
Nikhil Mishra*, Mostafa Rohaninejad*, Xi Chen, Pieter Abbeel.
In the proceedings of the International Conference on Learning Representations (ICLR), 2018.
Preliminary versions appeared at the NeurIPS 2017 Workshop on Meta-Learning and ICML 2017 Lifelong Learning: A Reinforcment Learning Approach workshop.

Prediction and Control with Temporal Segment Models (arxiv)
Nikhil Mishra, Pieter Abbeel, Igor Mordatch.
In the proceedings of the International Conference on Machine Learning (ICML), 2017.

Combining Model-Based Policy Search with Online Model Learning for Control of Physical Humanoids (pdf)
Igor Mordatch, Nikhil Mishra, Clemens Eppner, Pieter Abbeel.
In the proceedings of the International Conference on Robotics and Automation (ICRA), 2016.