Voiceprint extraction changelog

2.0.1 (2026-01-13)

Fixed

• [PLAT-1506] Fixed a rare issue where audio resampling could fail when processing large files, particularly when converting between certain sample rates (e.g., 11025 Hz to 8000 Hz).
• [PLAT-1509] Fixed GPU processing failure on NVIDIA Blackwell architecture GPUs (e.g., RTX 1000 series) caused by a bug in ONNX Runtime.
• Fixed license validation error handling for legacy (v1) license files with missing required fields. Previously, this could cause undefined behavior.

2.0.0 (2025-11-19)

Breaking change

• [PLAT-1425] Improve compatibility with future models - WARNING microservice requires at least model v5.4.0

Changed

• [PLAT-1471] Refactoring of the internal audio handling to reduce memory usage. Audio format support is now limited to WAV and FLAC.

1.7.0 (2025-10-09)

Changed

• Microservice now by default logs each request with incremental correlation id

Fixed

• [VOX-2666] Inconsistent logging
• [PLAT-1277] Possible memory leak on cuda device
• [PLAT-1320] Fix resource exhausted exception handling

Added

• [PLAT-1408] Added support for extracting vector voiceprints that can be used in vector databases.

1.6.0 (2024-04-28)

Added

• [PLAT-1056] Support for new licenses tied only to model major version. Old licenses tied to precise model version are still supported.
• [PLAT-1070] Default port 8080 exposed in Dockerfile
• [VOX-1293] phonexia.grpc.common.Status service
• [VOX-1293] Field license_flags to LicensingInfoResult in phonexia.grpc.common.Licensing

1.5.3 (2024-12-10)

Fixed

• [PLAT-978] Loading configuarion from model

1.5.2 (2024-10-03)

Fixed

• [PLAT-947] Missing libraries in GPU Docker image

1.5.1 (2024-10-01)

Fixed

• [PLAT-936] Crash when using negative values in time range

1.5.0 (2024-09-16)

Added

• [PLAT-888] Added PHX_KEEPALIVE_TIME_S and PHX_KEEPALIVE_TIMEOUT_S options for connection timeout specification

Changed

• Microservice argument device_id (Env. PHX_DEVICE_ID) to device_index (Env. PHX_DEVICE_INDEX)

1.4.0 (2024-07-10)

Added

• Performance optimizations for GPU image
• [PLAT-771] Support for RAW audio and real-time audio processing during transmission using the RAW audio
• [PLAT-781] Support for license extensions

Fixed

• [PLAT-758] Device is not logged by default

1.3.1 (2024-06-06)

Fixed

• [PLAT-788] Missing entrypoint in GPU docker image

1.3.0 (2024-05-30)

Added

• [PLAT-418] Reduce file size of voiceprints in UBJSON format.
• [PLAT-669] Add support for custom log tags in metadata and logs.
• [PLAT-774] Optimized docker images.

Fixed

• [PLAT-753] GPU docker image may fail to load under certain circumstances.

1.2.0 (2024-02-29)

Added

• [PLAT-448] Parameter to set the maximum amount of speech for extraction (config.speech_length)
• [PLAT-43] GPU enabled docker image for GPU acceleration of certain parts of the extraction algorithm

1.1.0 (2024-02-20)

Added

• [PLAT-463] Parameters for selecting the part of the audio to process in ExtractRequest (audio.time_range)
• [PLAT-603] Premature processing cancellation if connection is canceled by client
• [PLAT-625] Additional logging during service initialisation

Fixed

• [PLAT-614] Inconsistent license logging

1.0.0 (2023-12-13)

Changed

• [PLAT-314] Refactored protobuf messages
• [PLAT-313] Only mono channel recordings are supported
• [PLAT-203] Disable CPU Memory Arena in ONNXRuntime to avoid extensive memory pre-allocations

Fixed

• [PLAT-194] Empty recordings could cause error during processing
• [PLAT-207] Extractor crashes on older CPU architectures with SIGILL

Added

• [PLAT-520] Correlation ID for task tracking
• [PLAT-44] Implemented chunk processing for improved memory consumption during data processing
• [PLAT-357] Use status builder for error handling
• [PLAT-298] Endpoint for license checking
• [PLAT-249] Integrate license manager in microservice
• [PLAT-190] Use hybrid VAD (composition of NN, energy and spectrum VAD)