litert-community/pp-ocrv5-image-to-text

Image → text with PP-OCRv5

Choose an image, take one with your camera, drop or paste one, or try the bundled sample, and PP-OCRv5 finds each line of text and reads it; hover or tap a line to see where it is in the image. It runs in your browser on LiteRT.js: the detector on WebGPU when it is available and on WebAssembly otherwise, the recognizer on WebAssembly. Your image never leaves the page; the first visit downloads about 43 MB of model files from Hugging Face and caches them.

• —Source: samples/web_demos in google-ai-edge/litert-samples
• —Model: litert-community/PP-OCRv5-LiteRT