Scholar
Aleksei Petrenko
Google Scholar ID: G2zXCNkAAAAJ
Research Scientist at Apple, Computer Science PhD from the University of Southern California
Citations & Impact
Citations
536
H-index
10
i10-index
10
Publications
13
Co-authors
19
list available
Publications
Resume (English only)
Academic Achievements
- 2023: Co-authored 'DexPBT: Scaling up Dexterous Manipulation for Hand-Arm Systems with Population Based Training' (submitted to RSS 2023).
- 2023: Co-authored 'DeXtreme: Transfer of Agile In-hand Manipulation from Simulation to Reality' (ICRA 2023).
- 2021: Published 'Megaverse: Simulating Embodied Agents at One Million Experiences per Second' (ICML 2021), introducing the fastest embodied simulator at the time.
- 2021: Co-authored work on end-to-end deep RL for quadrotor swarms with sim-to-real transfer (CORL 2021).
- 2021: Co-authored 'Agents that Listen: High-Throughput RL with Multiple Sensory Systems' (IEEE Conference on Games 2021).
- 2021: Co-authored 'Large Batch Simulation for Deep Reinforcement Learning' (ICLR 2021).
- 2020: Published 'Sample Factory: Egocentric 3D Control from Pixels at 100000 FPS with Asynchronous RL' (ICML 2020), achieving state-of-the-art throughput.
- Led or contributed to open-source projects including Sample Factory and Megaverse.
