act_ImproveDatasetMerge is an open-source imitation-learning model that predicts short action chunks instead of single steps, trained on teleoperated data. It is designed for robotics and AI researchers to automate and improve action sequence prediction. The model supports training, evaluation, and fine-tuning for custom tasks.
In the Other AI space, Act ImproveDatasetMerge takes a focused approach. It focuses on automating the prediction of action sequences from teleoperated data for robotics and imitation learning tasks. Act ImproveDatasetMerge is an open-source project aimed at AI researchers and robotics developers. The project is open source (Apache-2.0). Act ImproveDatasetMerge is available on the web and API.
HaiqalQ19 builds and maintains Act ImproveDatasetMerge, and the product first shipped in 2024. The project is developed in the open on GitHub with 25.8k stars and 157 commits in the last 90 days. Among its 5 catalogued features are imitation learning, action chunk prediction, and teleoperated data training.
Latest indexed changes and source events
HaiqalQ19/act_ImproveDatasetMerge verified by the PulseGate indexer
Other apps tracked under the same category.
