A bimanual ACT policy runs at 10 Hz on an 8 GB Jetson Orin Nano Super with roughly 90-95% task success, and the paper documents when quantization is necessary and which layers TensorRT refuses to quantize.
Active vision might be all you need: Exploring active vision in bimanual robotic manipulation,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
background 1
citation-polarity summary
fields
cs.RO 1years
2026 1verdicts
CONDITIONAL 1roles
background 1polarities
support 1representative citing papers
citing papers explorer
-
Bimanual Manipulation Within an 8 GB Budget: Zero-Copy Sensing and Quantized ACT on an Entry-Level Jetson
A bimanual ACT policy runs at 10 Hz on an 8 GB Jetson Orin Nano Super with roughly 90-95% task success, and the paper documents when quantization is necessary and which layers TensorRT refuses to quantize.