Searcharxiv⌕ Search
Home/Archive/arXiv

arXiv · 2205.08948

i-MYO: A Hybrid Prosthetic Hand Control System based on Eye-tracking, Augmented Reality and Myoelectric signal

Chunyuan Shi·Dapeng Yang·Jingdong Zhao·Li Jiang
View original source ↗Read source document ↗Export metadata ↓

Abstract

Dexterous prosthetic hands have better grasp performance than traditional ones. However, patients still find it difficult to use these hands without a suitable control system. A new hybrid myoelectric control system, termed i-MYO, is presented and evaluated to solve this problem. The core component of the i-MYO is a novel grasp-type switching interface based on eye-tracking and augmented reality (AR), termed i-GSI. With the i-GSI, the user can easily switch a grasp type (six total) for a prosthetic hand by gazing at a GazeButton. The i-GSI is implemented in an AR helmet and is integrated, as an individual module, into the i-MYO system. In the i-MYO system, the myoelectric signal was used to control hand opening /closing proportionally. The operation of the i-MYO was tested on nine healthy subjects who wore HIT-V hand on the forearm and manipulated objects in a reach-and-grasp task. It was also tested on one patient who had an inferior myoelectric signal and was required to control the HIT-V hand to grasp objects. Results showed that in 91.6% of the trials, inexperienced healthy subjects accomplished the task within 5.9 s, and most failed trials were caused by a lack of experience in fine grasping. In addition, in about 1.5% of trials, the subjects also successfully transferred the objects but with a non-optimal grasp type. In 97.0% of the trials, the subjects spent ~1.3 s switching the optimal grasp types. A higher success rate in grasp type (99.1%) for the untrained patient has been observed thanks to more trials conducted. In 98.7 % of trials, the patient only needed another 2 s to control the hand to grasp the object after switching to the optimal grasp type. The tests demonstrate the control capability of the new system in multi-DOF prosthetics, and all inexperienced subjects were able to master the operation of the i-MYO quickly within a few pieces of training and apply it easily.

Explore related subjects

cs.ROcs.HC

Keep this discovery

BibTeXRIS

Chunyuan Shi, Dapeng Yang, Jingdong Zhao, Li Jiang. 2022-05-18. i-MYO: A Hybrid Prosthetic Hand Control System based on Eye-tracking, Augmented Reality and Myoelectric signal. https://arxiv.org/abs/2205.08948

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related papers

arXiv2026

IMLE-VLA: Fast Single-Step Action Generation for Vision-Language-Action Policies

Vision-language-action (VLA) policies leverage pretrained vision-language backbones to achieve strong cross-task generalization. A leading design couples this backbone with a dedicated continuous action head trained via diffusion or flow matching. However, such heads rely on iterative multi-step sampling, for example 10 Euler steps in $\pi_{0.5}$. This creates an inference bottleneck that produces stop-and-go movement in the robot and slower task completion. We introduce IMLE-VLA, which replaces the iterative action head with a single-step conditional generator trained via conditional Implicit Maximum Likelihood Estimation (cIMLE). The cIMLE objective promotes multimodal action coverage, avoiding the mode collapse of naive regression heads while eliminating multi-step sampling entirely. When IMLE-VLA is applied to $\pi_{0.5}$, it increases inference frequency 3.67x (55 Hz vs. 15 Hz), enabling up to 11x higher action throughput. On the 40-task LIBERO benchmark, IMLE-VLA achieves the highest average success rate (98.0%) among all baselines while leading in inference frequency. Under the test-time perturbations of LIBERO-plus, IMLE-VLA retains $\pi_{0.5}$'s robustness while other baselines degrade sharply, confirming that the cIMLE head preserves generalization. Real-world experiments on a Franka Emika Panda across four tasks demonstrate smoother motion (2.2x to 3.0x lower jerk) and faster task completion, with IMLE-VLA outperforming $\pi_{0.5}$ on every task and reducing average VLA inference time per episode by 3.9x to 6.6x. Videos and code are available at https://kianhk6.github.io/IMLE-VLA/

cs.RO↗
arXiv2026

ObstaDiff: Generalizable Diffusion Policy Learning via Obstacle-aware Representations

Imitation learning has achieved impressive results in robotic manipulation, yet most existing approaches assume clean backgrounds and lack explicit mechanisms for obstacle-aware motion generation. Extending such policies to cluttered, real-world scenes with unstructured obstacles remains a key generalization challenge. We present ObstaDiff, a decomposed diffusion-policy framework with a lightweight obstacle-aware visual encoder. ObstaDiff extracts a structured target-obstacle-background representation, enabling the downstream alignment policy to generate end-effector trajectories toward a target-centered bottleneck pose while reasoning about surrounding obstacles. We evaluate ObstaDiff on 61 real-robot greenhouse trials per method (366 executions in total). ObstaDiff achieves 75.41% average task success and 8.20% average obstacle collision rate, outperforming representative imitation-learning baselines and improving generalization in cluttered agricultural scenes.

cs.RO↗
arXiv2026

Testing Between the Test Cases: Proving End-to-End Steering in Conditions You Never Drove

AI-based automated vehicle testing is challenging because a model that passes every test condition can still fail in the real world. Formal verification offers a way to directly address this gap. On a simulated highway and an arterial road we trained two small end-to-end steering networks each in CARLA, one on clear conditions alone and one on clear, fog, night and low sun. All four models were driven against a 2.19 ft lane-departure budget. Without driving again, we used bound propagation, a formal method that reads the trained weights, to compute how far steering can drift at every disturbance strength between two captured images. One calculation covers more than a campaign could drive: on the arterial it spans 133 poses, where ten intensities each would be 10^133 combinations, in minutes on one GPU. Not only did formal verification find conditions that broke the clear-trained policy without simulation testing, it provided some preliminary evidence for potential failures between the test cases. Our overall conclusion is that formal verification is a viable complement to simulation, and could be adopted as a part of verification and validation for automated driving.

cs.RO↗