{"id":50727690,"url":"https://github.com/kernelshreyak/gesture-controlled-arm","last_synced_at":"2026-06-10T05:32:15.125Z","repository":{"id":346711194,"uuid":"1167546155","full_name":"kernelshreyak/gesture-controlled-arm","owner":"kernelshreyak","description":"Gestured controlled humanoid arm, precise movement controlled via hand pose detection(mediapipe HandLandmarker)","archived":false,"fork":false,"pushed_at":"2026-03-25T04:17:46.000Z","size":5762,"stargazers_count":0,"open_issues_count":0,"forks_count":0,"subscribers_count":0,"default_branch":"main","last_synced_at":"2026-03-26T09:51:42.670Z","etag":null,"topics":["bionic-arm","gazebo","humanoid-arm","humanoid-robot","mujoco-python","robotics","ros2","ros2-humble","webots"],"latest_commit_sha":null,"homepage":"","language":"Python","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":null,"status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/kernelshreyak.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":null,"code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null,"zenodo":null,"notice":null,"maintainers":null,"copyright":null,"agents":null,"dco":null,"cla":null}},"created_at":"2026-02-26T12:14:17.000Z","updated_at":"2026-03-25T04:19:56.000Z","dependencies_parsed_at":null,"dependency_job_id":null,"html_url":"https://github.com/kernelshreyak/gesture-controlled-arm","commit_stats":null,"previous_names":["kernelshreyak/gesture-controlled-arm"],"tags_count":null,"template":false,"template_full_name":null,"purl":"pkg:github/kernelshreyak/gesture-controlled-arm","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/kernelshreyak%2Fgesture-controlled-arm","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/kernelshreyak%2Fgesture-controlled-arm/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/kernelshreyak%2Fgesture-controlled-arm/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/kernelshreyak%2Fgesture-controlled-arm/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/kernelshreyak","download_url":"https://codeload.github.com/kernelshreyak/gesture-controlled-arm/tar.gz/refs/heads/main","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/kernelshreyak%2Fgesture-controlled-arm/sbom","scorecard":null,"host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":286080680,"owners_count":34139178,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2026-05-26T15:22:16.424Z","status":"online","status_checked_at":"2026-06-10T02:00:07.152Z","response_time":89,"last_error":null,"robots_txt_status":"success","robots_txt_updated_at":"2025-07-24T06:49:26.215Z","robots_txt_url":"https://github.com/robots.txt","online":true,"can_crawl_api":true,"host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["bionic-arm","gazebo","humanoid-arm","humanoid-robot","mujoco-python","robotics","ros2","ros2-humble","webots"],"created_at":"2026-06-10T05:32:11.046Z","updated_at":"2026-06-10T05:32:15.114Z","avatar_url":"https://github.com/kernelshreyak.png","language":"Python","funding_links":[],"categories":[],"sub_categories":[],"readme":"# Gesture-Controlled Bionic Arm\n\nThis repository is for building a camera-driven bionic arm system controlled by real-time hand tracking. The input side stays consistent across implementations: camera frames are processed with MediaPipe, landmarks are extracted, and those landmarks are mapped into a simulator or robot representation.\n\nThe current repository has two active directions:\n- a Python hand-tracking and simulation pipeline\n- a Webots project for the bionic arm model and world\n\nFuture simulator backends such as MuJoCo and Genesis will be added later, but they will reuse the same tracker stage: camera input followed by MediaPipe hand landmark detection.\n\n## Core Flow\n\n1. Capture video from the configured camera device.\n2. Run MediaPipe HandLandmarker on each frame.\n3. Convert landmarks into joint or pose targets.\n4. Drive a simulator or robot model from those targets.\n\n## Main Scripts\n\n- `run_pipeline.py`\n  Main Python entry point for the current working pipeline. It reads `config.yaml`, runs the hand tracker, maps landmarks into joint targets, and drives the simplified PyBullet hand.\n\n- `hand_pose_detection_test.py`\n  Standalone tracker visualization script. Use this to verify camera input, MediaPipe detection, and on-screen hand skeleton rendering before debugging simulator behavior.\n\n## Python Modules\n\n- `src/hand_tracker.py`\n  MediaPipe hand tracking wrapper. Handles camera capture and returns 2D and world landmarks.\n\n- `src/mapping.py`\n  Landmark conversion and finger curl mapping logic used by the simulation pipeline.\n\n- `src/sim_pybullet.py`\n  Simplified CPU-friendly hand simulation backend built on PyBullet.\n\n- `src/visualization.py`\n  Shared drawing helpers for rendering the hand skeleton on video frames.\n\n- `src/config.py`\n  YAML config loader used by the scripts.\n\n## Webots Project\n\nThe Webots project lives under `bionic_arm_webots/`.\n\n- `bionic_arm_webots/worlds/bionic_arm.wbt`\n  Main Webots world for the arm model.\n\n- `bionic_arm_webots/CAD/simplified_bionic_arm.FCStd`\n  CAD source for the simplified arm model.\n\nThe Webots side is currently separate from the Python simulation pipeline. It is the main place for arm/world modeling and scene iteration.\n\n## Configuration\n\nAll runtime configuration is in `config.yaml`.\n\nImportant values include:\n- camera device path\n- frame size\n- MediaPipe model path\n- tracker confidence thresholds\n- simulation timing and motor settings\n- landmark-to-simulation scaling\n\nDefault camera device:\n- `/dev/video2`\n\n## Models\n\nMediaPipe task models live in `models/`.\n\n- `models/hand_landmarker.task`\n  Required for the current tracker and pipeline scripts.\n\nSee [models/README.md](/home/shreyak/programming/gesture-controlled-arm/models/README.md) for model placement notes.\n\n## Setup\n\nInstall dependencies:\n\n```bash\npip install -r requirements.txt\n```\n\nRun the main pipeline:\n\n```bash\npython run_pipeline.py\n```\n\nRun the tracker-only test window:\n\n```bash\npython hand_pose_detection_test.py\n```\n\n## Notes\n\n- The Python pipeline currently uses PyBullet because it is lightweight and easy to iterate on from Python.\n- MuJoCo and Genesis-based implementations can be added later without changing the tracker stage.\n- The repository is still evolving, so interfaces and file layout may continue to shift as the simulator backends mature.\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fkernelshreyak%2Fgesture-controlled-arm","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fkernelshreyak%2Fgesture-controlled-arm","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fkernelshreyak%2Fgesture-controlled-arm/lists"}