A DearPyGui-based application that scans a physical Rubik's Cube with a webcam using YOLO for detection, reconstructs face colors, and produces a solution using Kociemba's algorithm.

• Python 3.9+
• Webcam (for scanning)
• Recommended GPU support for PyTorch (falls back to CPU)
• Python packages: in requirements.txt

Install the dependencies (PowerShell example):

python -m pip install -r requirements.txt

Adjust the PyTorch install command for your CUDA version or use the CPU-only wheel if needed.

• main.py — main GUI scanner/solver; runs the webcam pipeline and visualization.
• rubik_face_read.py — cube state representation, move application, and color extraction helpers.
• YOLO.pt — pretrained YOLO model for cube detection.

From the repository root:

python main.py

Controls inside the app:

• SPACE — capture the current face
• R — redo the last captured face
• C — clear all captured faces
• V — open 3D view
• A — open solution animator
• Arrow keys — rotate the 3D/animation view
• Q — quit

Workflow:

1. Align a cube face in the camera view (follow the order) and press SPACE to capture each face. Guide:

• White (on top is blue) -> Yellow (on top is green) -> Green (on top is white) -> Orange (on top is white) -> Blue (on top is white) -> Red (on top is white).

2. The app resolves colors, validates the state, and runs Kociemba to compute a solution string.
3. Use V to inspect the reconstructed cube in 3D.
4. Use A to play/pause/step through the solution animation.

• Webcam not opening: ensure no other app is using the camera; check the cv2.VideoCapture(0) index in main.py.
• Slow inference: confirm PyTorch is using CUDA; otherwise it will run on CPU.
• Color misreads: verify lighting and try recapturing faces; adjust thresholds in rubik_face_read.py if needed.
• Kociemba install error: This is a classic error when programming Python on Windows. Cause: The kociemba library (Rubik solving algorithm) is written in C for speed. When you use pip install, it tries to compile this C source code on your machine, but your machine lacks the compiler toolset named Microsoft Visual C++ Build Tools.

MIT License (see LICENSE).