Lines Matching full:inference
16 …ic Speech Recognition pipeline with 3 steps: data pre-processing, inference execution and inference
26 * @param executor - unique pointer to inference runner
27 * @param decoder - unique pointer to inference results decoder
35 * Preprocesses and prepares the data for inference by
51 * @brief Executes inference
53 * Calls inference runner provided during instance construction.
55 …* @param[in] preprocessedData - input inference data. Data type should be aligned with input tenso…
56 * @param[out] result - raw inference results.
59 … void Inference(const std::vector<T>& preprocessedData, common::InferenceResults<int8_t>& result) in Inference() function in asr::ASRPipeline
66 * @brief Standard inference results post-processing implementation.
68 * Decodes inference results using decoder provided during construction.
70 * @param[in] inferenceResult - inference results to be decoded.