Add dataset download link to README

Final round of README polish: restore assets, add Why This Project section, add architecture image, improve modeling text, and update Author info

Incorporate comprehensive README review feedback: restructure sections, refine metrics phrasing, simplify repository tree, and re-add YAML frontmatter

Fix video embedding by using raw URL for GitHub auto-rendering

Move Demo.mp4 to assets folder and update README video embed

Use exact user provided images for README UI demo and samples

Update README appearance with demo UI, sample images, demo video, and test_images instruction

Update docs, readme, and add project context

Remove Developer button from navbar

Upgrade python base image to 3.12-slim to satisfy numpy 2.5.0 requirements

Replace obsolete libgl1-mesa-glx dependency with libgl1

Fix colorTo value in Hugging Face metadata

Add Hugging Face Space metadata header to README.md

Support single-container Docker build with static frontend export for Hugging Face Spaces

Fix backend tests: resolve hardcoded venv python path and relative test image path

Update repository references from captcha-ocr to VisionSeq

Restructure repository: modularize backend and frontend, add docs, Dockerfiles, and notebooks

Swap image positions in README

Add live demo link and fix screenshot references in README

Update PyTorch to use CPU version for Streamlit Cloud

Add final ResNet18 model

Initial commit for CAPTCHA OCR (excluding large .pth files)