onnx-runtime-web

🎨 Free AI watermark remover for images & videos. Runs entirely in your browser — no uploads, no accounts, no cloud. Open source.

Client Side Voice Changer is a browser-only voice conversion app built with Next.js. It records or uploads speech, transcribes it locally, synthesizes speech from the transcript, applies lightweight voice effects, and lets the user preview and download the result as a WAV file. Achieved via huggingface-transformers and kokoro

A 36M-parameter goldfish language model with a 10-second memory + pixel-art PWA desk pet. Runs in your browser, fully offline. Adopt it at den-sec.github.io/glublm/desk-pet/

스마트폰 카메라로 서가를 읽는 온디바이스 AI — 청구기호 판독·오배열 감지·도서찾기 (판독 99% 실측, 서버 0원)

Seven end-to-end CNN and computer-vision projects covering image segmentation, object detection, medical imaging, transfer learning, DenseNet, ResNet, VGG16, AlexNet-style CNNs, MobileNetV2, Grad-CAM, TensorFlow.js, ONNX Runtime Web, Vercel, and GitHub Pages.

Ten end-to-end Transformer projects covering NLP, retrieval, long-context QA, LoRA/PEFT instruction tuning, multimodal AI, Vision Transformers, CLIP, ONNX browser inference, and RAG—deployed through Hugging Face Spaces, GitHub Pages, and Vercel.

AI CCTV Threat Monitor - 3-model YOLOv8 ensemble with threat levels (loitering, vehicle lurking, weapons, altercations). FastAPI dashboard + in-browser ONNX demo on Vercel.

⚡ Serverless image+audio+text inference with Gemma 2, running 100% on-device via Chrome WebGPU. Zero backend GPU, $0.00/month.

Real-time Person Re-Identification with Body Part Segmentation

PP-DocLayoutV3 browser SDK powered by ONNX Runtime Web, with WebGPU/WASM, FP16/FP32, workers, caching, and custom model manifests.