Author

Keita Ota

h-index 10 View on Semantic Scholar

We have 3 of 37 papers

We haven’t gathered this author’s papers yet. Follow them and we’ll fetch their work.

Not the right person? Other researchers publish under this name.

Preprint Oct 2026

RoboPace: Contact-Aware Time-Optimal Retiming for Action-Chunk Policies

Robot manipulation data collection has been shifting from teleoperation toward robot-free demonstrations, through interfaces such as the Universal Manipulation Interface (UMI) or directly from human hands. Vision-Language-Action (VLA) policies trained on such data inherit the demonstrator's timing. Yet human timing doe...

Mimo Shirasaka, Takehiko Ohkawa, Takuya Okubo et al. · 0 citations

Open access Aug 2026

Auditing Instruction–Trajectory Mismatches in Multimodal Robot Demonstrations

Multimodal Probabilistic Fusion (MMPF), a training-free auditing framework that treats each modality as an expert, estimates a task-label distribution from local neighborhood agreement and global prototype similarity, and then fuses modalities with predictive-entropy weighting in a product of experts is proposed.

Simon Holk, Ryosuke Takanami, T. Matsushima et al. · IEEE Robotics and Automation... · 0 citations

Preprint Aug 2026

CounterAlign: Counterfactual Supervision for Vision-Language-Action Models

This work synthesizes counterfactual instruction-observation-action tuples from the dataset and combines them with adversarial discriminator training to learn an instruction-grounded reward model for offline RL, without collecting additional rollouts or annotations.

Haruo Kondoh, Keita Ota, Asako Kanezaki et al. · 1 citation