tokimo-lab/tokimo-perception
On-device perception stack for Rust: OCR, CLIP, face recognition & streaming STT — one crate, many ONNX models, with lazy loading, 3-minute idle eviction, and an optional sidecar worker that exits on idle.
GitHub repository with 5 stars and 0 forks.
Language: C
Topics: clip, face-recognition, machine-learning, ocr, onnx, onnxruntime, rust, sherpa-onnx, speech-to-text