Keyword Detection (Live Audio + ML) — Rust

Source: apps/rust/heap/example_keyword_live/src/lib.rs (also apps/rust/zeroheap/example_keyword_live/src/lib.rs) | WASM demo not available

Real-time 'yes'/'no' keyword detection using on-board microphones and the TensorFlow Lite Micro micro_speech model. Audio is captured via I2S DMA, spectral features are extracted, and a neural network classifier runs inference every ~1 second.

Language-specific patterns

This example uses the ove Rust crate with no_std support. Error handling uses Result<T, ove::Error> with per-op narrow error sets. Shared state uses atomics; cross-thread kernel-object sharing uses Arc<T> in heap mode (from ove::heap::Arc) and ove::shared! cells (InitCell<T>) in zero-heap mode. #[ove::main] exports the C-ABI entry symbol.

See the overview for architecture details and the full API list.

How to build

# Native POSIX
make host.posix.example_keyword_live_rust
make configure && make download && make && make run