Online mode

.github/workflows/frontend-ng-ci.yml

@@ -0,0 +1,31 @@
+name: frontend-ng-ci
+
+on:
+  pull_request:
+    paths:
+      - 'frontend-ng/data-engine/**'
+      - '.github/workflows/frontend-ng-ci.yml'
+  push:
+    branches: [main]
+    paths:
+      - 'frontend-ng/data-engine/**'
+      - '.github/workflows/frontend-ng-ci.yml'
+
+jobs:
+  frontend-ng:
+    runs-on: ubuntu-latest
+    defaults:
+      run:
+        working-directory: frontend-ng/data-engine
+    steps:
+      - uses: actions/checkout@v4
+      - uses: actions/setup-node@v4
+        with:
+          node-version: 22
+          cache: npm
+          cache-dependency-path: frontend-ng/data-engine/package-lock.json
+      - run: npm ci
+      - run: npm run build
+      - run: npm run test -- --watch=false
+      - run: npm run lint
+      - run: npm run format:check

.gitignore

@@ -1,7 +1,23 @@
 __pycache__/
-*masks/
-video/
-bedroom/
+*.py[cod]
+.venv/
 build/
+*.log
+/bedroom/
+/video/
+!/backend/video/
+!/backend/video/**
+/backend/video/__pycache__/
+/.sam2_frame_cache/
+/backend/.data_engine_frames/
+/backend/.data_engine_windows/
+/backend/cache/
+/backend/saved/
+/backend/tests/bedroom/
+/backend/tests/apple/
+*masks/
+*_tracked_*.mp4
+*_tracking_*.json
 apple_*
-*.log
+!apple.mp4
+.DS_Store

.gitmodules

@@ -0,0 +1,3 @@
+[submodule "sam2"]
+	path = sam2
+	url = https://github.com/facebookresearch/sam2.git

README.md

@@ -2,30 +2,56 @@
 
 A complete data annotation and tracking system with both Python (FastAPI) backend and C++ interface. Supports video object segmentation using SAM 2 and point tracking using CoTracker.
 
+Use `git clone --recursive https://github.com/Rumarino-Team/data-engine.git` when cloning the repo
+
 ## Install Dependencies
 
-1. Create Python virtual environment using venv or Anaconda
+1. Create Python virtual environment using venv or Anaconda (mainly written and tested on Python 3.12.12)
+2. If you didn't use `--recursive` when doing git clone, then run `git submodule update --init --recursive` to make sure `sam2` is downloaded
+3. Run `pip install -e .` while in `sam2` directory
+4. Run `pip install -e ".[notebooks]"` while in `sam2` directory
+5. Return to the data engine directory and run `pip install -r requirements.txt`
 
-2. Run `git clone https://github.com/facebookresearch/sam2.git && cd sam2` outside of this directory
+## Using the Backend
+1. source the venv or Anaconda (depending on your shell)
+2. Run `fastapi dev backend/api.py` to use the API manually with hot reload, Run `fastapi run backend/api.py` to use the static API
 
-3. Run `pip install -e .` while in `sam2` directory
+Add the `--host xxx.xxx.xxx.xxx` and/or `--port xxxx` args to change its ip and port
 
-4. Run `pip install -e ".[notebooks]"` while in `sam2` directory
+## Testing the Python Backend
+1. Run `python3 backend/tests/tester.py` to test the API automatically and see results under `backend/tests/`
 
-5. Return to the data engine directory and run `pip install -r requirements.txt`
+## Angular Web Frontend
+Mainly for developing the frontend, for production its recommended to use the Tauri Desktop env instead
 
-## Testing via Python
+1. Change into `frontend-ng/data-engine`
+2. Install dependencies with `bun install` or `npm install`
+3. Start the frontend with `bun run start` or `npm run start`
+4. Keep the Python backend running separately
 
-1. Run `fastapi dev backend/api.py` to test the API manually with hot reload, or run `python3 backend/tests/tester.py` to test the API automatically and see results under `backend/tests/`
+The frontend defaults to `http://127.0.0.1:8000` for backend requests and also exposes a compact API URL override in the top bar.
+
+## Tauri Desktop Frontend
+
+1. Change into `frontend-ng/data-engine`
+2. Install JavaScript dependencies with `bun install` or `npm install`
+3. Install the Rust toolchain and Tauri prerequisites for your OS
+4. Start the Python backend separately
+5. Run `bun run tauri:dev` or `npm run tauri:dev:npm` for the desktop dev workflow
+6. Run `bun run tauri:build` or `npm run tauri:build:npm` to produce a packaged desktop build
+
+The Tauri app only wraps the Angular frontend. It does not bundle or launch the Python backend.
 
 ## Testing via C++
+Currently this part is behind the python api's backend implementation so it isnt recommended to use
 
 1. Install `nlohmann_json` and `curl` dev packages via your package manager
-
+2. Change into `/cpp-backend-bindings`
 2. Run `mkdir build`
-
 3. Run `cmake build` to generate CMake build files
-
 4. Run `cmake --build build` to compile project
+5. Run `./build/src/BackendInterfaceTests` while the venv is sourced to test backend code
+
 
-5. Run `./build/src/BackendInterfaceTests` while the venv is sourced to test backend code
+## Structure
+TBA

src/tests.cpp → backend-interface-cpp/src/tests.cpp

@@ -12,7 +12,7 @@ const std::string BASE_URL = "http://localhost:8000";
 const std::string BEDROOM_ZIP_URL =
     "https://dl.fbaipublicfiles.com/segment_anything_2/assets/bedroom.zip";
 const std::string BEDROOM_DIR = "tests/bedroom";
-const std::string TRACKING_VIDEO_PATH = "apple.mp4";  // relative to repo root
+const std::string TRACKING_VIDEO_PATH = "backend/tests/apple/apple.mp4";  // relative to repo root
 
 void runVideoTests(BackendInterface& backend) {
   std::cout