[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"project-1533":3},{"id":4,"name":5,"fullName":6,"owner":7,"repo":5,"description":8,"homepage":9,"htmlUrl":9,"language":10,"languages":9,"totalLinesOfCode":9,"stars":11,"forks":12,"watchers":13,"openIssues":14,"contributorsCount":15,"subscribersCount":15,"size":15,"stars1d":15,"stars7d":15,"stars30d":15,"stars90d":15,"forks30d":15,"starsTrendScore":15,"compositeScore":16,"rankGlobal":9,"rankLanguage":9,"license":17,"archived":18,"fork":18,"defaultBranch":19,"hasWiki":18,"hasPages":18,"topics":20,"createdAt":9,"pushedAt":9,"updatedAt":29,"readmeContent":30,"aiSummary":31,"trendingCount":15,"starSnapshotCount":15,"syncStatus":32,"lastSyncTime":33,"discoverSource":34},1533,"chek-ego-miner","chekdata\u002Fchek-ego-miner","chekdata","Crowdsource EGO robot data capture, contribution, and public-safe edge-host bring-up.",null,"Rust",201,13,5,4,0,3.44,"Apache License 2.0",false,"main",[21,22,23,24,25,26,27,28],"ai-agents","computer-vision","crowdsourcing","dataset","edge-computing","humanoid","ios","robotics","2026-06-12 02:00:29","[English](.\u002FREADME.md) | [简体中文](.\u002FREADME.zh-CN.md)\n\n# CHEK EGO Miner\n\nCapture first-person EGO data with a phone and computer, contribute sessions,\nand browse reusable datasets.\n\n## Start Here\n\n- Download the iOS app: [TestFlight](https:\u002F\u002Ftestflight.apple.com\u002Fjoin\u002FRrYdeDUv)\n- Choose your hardware: [Hardware Guide](.\u002Fdocs\u002Fhardware.md)\n- Check the current public roadmap: [TODO](.\u002FTODO.md)\n- See validation status: [Public Validation Matrix](.\u002Fdocs\u002Fpublic-validation-matrix.md)\n- Get step-by-step help:\n  - [Codex Guide](.\u002Fdocs\u002Fagent-guides\u002Fcodex.md)\n  - [Claude Guide](.\u002Fdocs\u002Fagent-guides\u002Fclaude.md)\n  - [OpenClaw Guide](.\u002Fdocs\u002Fagent-guides\u002Fopenclaw.md)\n- Browse and download contributed datasets:\n  - [EGO Dataset Portal](https:\u002F\u002Fwww.chekkk.com\u002Fhumanoid\u002Fego-dataset)\n\n## What You Can Do\n\n- start with one phone and one computer\n- add a stereo camera when you want better spatial cues\n- move to a dedicated edge setup for higher-throughput capture\n- use an AI assistant for guided install and troubleshooting\n- contribute sessions and explore downloadable datasets\n\n## Public-First Scope\n\nThis repository is the public-first entry point for contributors. It should\ngive people a clear install path, working operator surfaces, agent-guided\ntroubleshooting, and a usable frontend without forcing them to understand the\nfull internal runtime topology first.\n\nIt is the path for people who want to assemble their own edge machine or run\nthe stack from a computer install, instead of depending on a factory-integrated\ndevice workflow.\n\nAs the project evolves, this repo should keep that public-first experience\nwhile sharing common building blocks with the factory edge engineering line.\nThe goal is to avoid long-term duplicate runtime or module implementations,\nwhile still supporting a genuinely different installation and hardware path.\n\n## Capability Lanes\n\nFrom the public product point of view, `chek-ego-miner` should expose usable\nentry points for `SLAM`, `VLM`, and `time-sync` so that a self-assembled edge\nhost or a computer install can actually run those capabilities.\n\nThat does not mean all three should become long-term `ego-miner`-only module\nimplementations:\n\n- `SLAM`\n  - today is still more tightly coupled to the factory-integrated edge line,\n    especially around sensing bring-up, calibration, replay, training gates,\n    and engineering observability\n  - `chek-ego-miner` should expose the public install and operator path for it,\n    but should not fork a second long-lived core SLAM stack\n- `VLM`\n  - must remain directly usable from `chek-ego-miner`, including model fetch,\n    sidecar startup, service wiring, and public diagnostics\n  - the underlying runtime behavior should converge with the factory edge line\n    instead of drifting into two different VLM implementations\n- `time-sync`\n  - is a shared capture-quality capability needed by both product lines\n  - `chek-ego-miner` should surface install, validation, and operator feedback,\n    while deeper factory calibration and engineering observability can stay in\n    `chek-edge-runtime`\n\nThe rule going forward is simple: if two files express the same capability in\n`modules\u002F`, `profiles\u002F`, `services\u002F`, install backends, or shared UI panels,\nthey should be deduped into shared building blocks, templates, or versioned\nassets instead of being maintained as two drifting copies.\n\n## System View\n\n```mermaid\nflowchart LR\n  Phone[\"iPhone + CHEK App\"] --> Host[\"Computer or Edge Host\"]\n  Camera[\"Your Camera or Stereo Camera\"] --> Host\n  Agent[\"Codex \u002F Claude \u002F OpenClaw\"] --> Host\n  Host --> Upload[\"Upload EGO Sessions\"]\n  Upload --> Portal[\"Dataset Portal\"]\n  Upload --> Rewards[\"Token Rewards\"]\n```\n\n## Choose a Setup\n\n| Tier | Setup | Who it is for |\n| --- | --- | --- |\n| `Lite` | computer + your own camera | fastest way to start |\n| `Stereo` | computer + stereo camera | better spatial quality |\n| `Pro` | edge machine + stereo camera | dedicated capture and higher throughput |\n\nYou will also need a first-person phone mount. See [Hardware Guide](.\u002Fdocs\u002Fhardware.md)\nfor buying criteria, setup tradeoffs, search keywords, and direct purchase\nexamples including China marketplace links.\n\n## Get Step-by-Step Help\n\nIf you want guided setup instead of reading long docs, start with:\n\n- [AGENTS.md](.\u002FAGENTS.md)\n- one of the ready-to-use prompts:\n  - [Lite Install Prompt](.\u002Fprompts\u002Finstall-lite.md)\n  - [Stereo Install Prompt](.\u002Fprompts\u002Finstall-stereo.md)\n  - [Pro Edge Install Prompt](.\u002Fprompts\u002Finstall-pro-edge.md)\n  - [Camera Troubleshooting Prompt](.\u002Fprompts\u002Ftroubleshoot-camera.md)\n\nRecommended flow:\n\n1. Tell the assistant which hardware tier you have.\n2. Share your OS and what is already installed.\n3. Ask for one step at a time.\n4. Keep hardware checks, app install, and camera validation in the flow.\n\n## Before You Install\n\nBefore a longer install session, run the lightweight host self-check:\n\n```bash\npython3 scripts\u002Fcheck_host_basics.py\n```\n\nIf you plan to share your own fork or public changes, run:\n\n```bash\n.\u002Fscripts\u002Fscan_public_safety.sh .\n```\n\nOr use the CLI:\n\n```bash\n.\u002Fcli\u002Fchek-ego-miner doctor\n.\u002Fcli\u002Fchek-ego-miner camera-probe\n.\u002Fcli\u002Fchek-ego-miner readiness --tier lite\n.\u002Fcli\u002Fchek-ego-miner readiness --tier pro\n```\n\nUse `.\u002Fcli\u002Fchek-ego-miner camera-probe --capture-smoke` when you need to\ndistinguish \"camera is listed by the OS\" from \"the current terminal session can\nopen the camera and read a frame\".\n\n## Lite Setup on Linux or macOS\n\nIf you want the quickest supported setup path, start here:\n\n```bash\n.\u002Fcli\u002Fchek-ego-miner install \\\n  --profile basic \\\n  --apply \\\n  --system-install \\\n  --enable-services\n\npython3 -m pip install --user --break-system-packages -r scripts\u002Fedge_phone_vision_requirements.txt\n.\u002Fcli\u002Fchek-ego-miner fetch-phone-vision-models --json\n.\u002Fscripts\u002Fstart_edge_phone_vision_service.sh\n\n.\u002Fcli\u002Fchek-ego-miner basic-e2e \\\n  --edge-base-url http:\u002F\u002F127.0.0.1:8080 \\\n  --edge-token chek-ego-miner-local-token \\\n  --trip-id trip-public-basic-e2e \\\n  --session-id sess-public-basic-e2e \\\n  --output-dir .\u002Fartifacts\u002Fbasic-e2e \\\n  --json\n```\n\nIf Homebrew-managed macOS `python3` blocks `pip install --user`, install the\nsame requirements into a compatible interpreter such as `python3.10`; the\nstart script will auto-select it when available.\n\nAfter the basic flow finishes, you should see:\n\n- `ok: true`\n- `validation.ok: true`\n- `validation.score_percent: 100.0`\n- `public_download\u002Fdemo_capture_bundle.json` in your output directory\n\nNotes:\n\n- This path is intended for `Linux x86_64` and `macOS arm64` basic hosts.\n- On macOS, `install --system-install --enable-services` stages the runtime\n  under `~\u002F.chek-edge\u002Fruntime\u002Fmacos\u002Fbasic`.\n- `time_sync_samples` can stay empty on the single-phone basic path.\n\n## Training Threshold Validation\n\nRaw upload\u002Fdownload success is not the same as training readiness. To check a\ndownloaded session bundle against the public SLAM + time-sync candidate gate:\n\n```bash\npython3 scripts\u002Fgenerate_slam_time_sync_benchmark.py \\\n  --bundle \u002Fpath\u002Fto\u002Fraw_bundle.tar.gz \\\n  --tier pro \\\n  --output \u002Ftmp\u002Fslam_time_sync_benchmark.json \\\n  --json\n\npython3 scripts\u002Fvalidate_training_thresholds.py \\\n  --bundle \u002Fpath\u002Fto\u002Fraw_bundle.tar.gz \\\n  --tier pro \\\n  --slam-benchmark-report \u002Ftmp\u002Fslam_time_sync_benchmark.json \\\n  --json\n```\n\nThe validator checks:\n\n- VLM events, segments, fallback usage, and latency\n- time-sync sample count, accepted mapping ratio, per-source RTT, and offset span\n- phone pose, stereo pose, Wi-Fi pose, and fisheye track completeness\n- whether a SLAM benchmark report with drift, reprojection, pose-graph, and\n  body-tracking metrics exists and passes the candidate budgets\n\nThe benchmark generator only emits metrics that can be computed from the bundle\nfacts. It can compute stereo reprojection error and body-tracking coverage from\nthe current raw bundle. It keeps trajectory drift and pose-graph residual as\nexplicit blockers until the bundle includes ground truth, loop-closure evidence,\nor a SLAM optimizer report.\n\nIt returns exit code `0` only when `training_ready=true`; incomplete or\ncandidate-only bundles return exit code `2` and list the blocking checks. Raw\n`clock_offset_ns` values can cross clock domains, so the validator uses per\nsource-kind offset span\u002Fstability instead of treating the absolute offset as the\nsync error.\n\nIt intentionally separates:\n\n- `signal_candidate_ready`: the bundle has enough live signals for a candidate\n  review\n- `training_ready`: the bundle passes frozen thresholds and has required SLAM\n  benchmark metrics\n\nUntil `configs\u002Fslam_time_sync_training_v1.json` is frozen and real benchmark\nmetrics are present, the tool will refuse to claim final training readiness.\n\n## Pro Setup on Jetson\n\nIf you want the full `Pro` runtime path on Jetson, bootstrap the machine first.\nThis brings in:\n\n- stereo calibration\n- the Wi-Fi sensing model and `sensing-server`\n- `edge-orchestrator`, `ruview-leap-bridge`, and `ruview-unitree-bridge` binaries\n- `RuView\u002Fui-react\u002Fdist`\n- an existing Jetson GPU VLM environment plus SmolVLM model cache\n\n```bash\n.\u002Fcli\u002Fchek-ego-miner jetson-professional-bootstrap -- --force\n.\u002Fcli\u002Fchek-ego-miner install \\\n  --profile professional \\\n  --apply \\\n  --system-install \\\n  --runtime-edge-root \"$PWD\"\n```\n\nIf you only want the Jetson VLM path, use the bundled sidecar and model fetch\nflow:\n\n```bash\n.\u002Fcli\u002Fchek-ego-miner install \\\n  --profile professional \\\n  --apply \\\n  --system-install \\\n  --enable-services\n\npython3 -m pip install --user -r scripts\u002Fedge_vlm_requirements.txt\n.\u002Fcli\u002Fchek-ego-miner fetch-vlm-models --json\n.\u002Fcli\u002Fchek-ego-miner vlm-start\n```\n\nIf the target Jetson already has a working GPU VLM environment and local model\ncache, you can wire only those VLM assets and enable the sidecar through\n`systemd-user`:\n\n```bash\n.\u002Fcli\u002Fchek-ego-miner jetson-vlm-bootstrap -- --force\n.\u002Fcli\u002Fchek-ego-miner service-install \\\n  --profile professional \\\n  --service chek-edge-vlm-sidecar \\\n  --enable \\\n  --runtime-edge-root \"$PWD\"\n```\n\nNotes:\n\n- `fetch-vlm-models` downloads the core Hugging Face files needed by\n  `transformers`.\n- Default model files are stored under `model-candidates\u002Fhuggingface\u002F`.\n- A successful Jetson bring-up should look like:\n  - `.\u002Fcli\u002Fchek-ego-miner readiness --tier pro` reports the host is ready\n  - required services reach `active`\n  - `\u002Fhealth`, `\u002Fassociation\u002Fhint`, `\u002Fapi\u002Fv1\u002Fstream\u002Fstatus`, and `\u002Finfer`\n    return live responses on the host\n\n## Dataset Portal\n\nSearch and download contributed data from:\n\n- [https:\u002F\u002Fwww.chekkk.com\u002Fhumanoid\u002Fego-dataset](https:\u002F\u002Fwww.chekkk.com\u002Fhumanoid\u002Fego-dataset)\n\n## What You Can Do Today\n\n- onboard a new capture setup\n- choose hardware and accessories\n- use prompts for guided setup\n- run the Lite\u002Fbasic path on Linux or macOS\n- bring up the Pro Jetson VLM and service path\n- learn how contribution, rewards, and dataset discovery work\n\n## Docs\n\n- [Public Roadmap \u002F TODO](.\u002FTODO.md)\n- [Public Validation Matrix](.\u002Fdocs\u002Fpublic-validation-matrix.md)\n- [Hardware Guide](.\u002Fdocs\u002Fhardware.md)\n- [Quickstart](.\u002Fdocs\u002Fquickstart.md)\n- [Hardware\u002FProfile Mapping](.\u002Fdocs\u002Fprofile-mapping.md)\n- [Diagnostics](.\u002Fdocs\u002Fdiagnostics.md)\n- [Token Rewards](.\u002Fdocs\u002Ftoken-rewards.md)\n- [Privacy, Consent, and Data License](.\u002Fdocs\u002Fprivacy-data-license.md)\n- [FAQ](.\u002Fdocs\u002Ffaq.md)\n- [Codex Guide](.\u002Fdocs\u002Fagent-guides\u002Fcodex.md)\n- [Claude Guide](.\u002Fdocs\u002Fagent-guides\u002Fclaude.md)\n- [OpenClaw Guide](.\u002Fdocs\u002Fagent-guides\u002Fopenclaw.md)\n\n## Contributing\n\nSee [CONTRIBUTING.md](.\u002FCONTRIBUTING.md).\n\n## Security\n\nSee [SECURITY.md](.\u002FSECURITY.md).\n\n## License\n\nSee [LICENSE](.\u002FLICENSE).\n","CHEK EGO Miner 是一个用于采集第一人称视角（EGO）数据、贡献会话并提供公共安全边缘计算环境的开源项目。其核心功能包括通过手机和电脑捕捉EGO数据，支持立体相机以获取更好的空间线索，并允许用户将数据上传至可重用的数据集。该项目采用Rust语言编写，具备良好的性能与安全性。它适合需要构建或扩展机器人视觉数据集的研究人员及开发者使用，特别是在人工智能代理、计算机视觉以及边缘计算等领域。此外，项目提供了详细的硬件指南、安装步骤以及AI助手指导安装和故障排除，确保用户能够顺利上手。",2,"2026-06-11 02:44:33","CREATED_QUERY"]