|
| 1 | +# Changelog |
| 2 | + |
| 3 | +All notable changes to this project will be documented in this file. |
| 4 | + |
| 5 | +The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.1.0/), |
| 6 | +and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html). |
| 7 | + |
| 8 | +## [Unreleased] |
| 9 | + |
| 10 | +## [0.2.14] - 2026-04-30 |
| 11 | + |
| 12 | +### Bugfixes |
| 13 | + |
| 14 | +- Allow classifiers trained with hidden units and with append mode(#33, #22) |
| 15 | + |
| 16 | +## [0.2.13] - 2026-04-06 |
| 17 | + |
| 18 | +### Changed |
| 19 | + |
| 20 | +- Changed sigmoid function to match birdnet_analyzer by @Josef-Haupt |
| 21 | + |
| 22 | +### Bugfixes |
| 23 | + |
| 24 | +- Fixed issue #29 |
| 25 | + |
| 26 | +## [0.2.12] - 2026-02-22 |
| 27 | + |
| 28 | +### Added |
| 29 | + |
| 30 | +- Added convenience functions to export embeddings by @Josef-Haupt |
| 31 | +- Added some code documentation |
| 32 | + |
| 33 | +### Changed |
| 34 | + |
| 35 | +- Updated flat sigmoid to match birdnet_analyzer by @Josef-Haupt |
| 36 | + |
| 37 | +### Bugfixes |
| 38 | + |
| 39 | +- Fixed issue with one test on Python 3.11 & 3.12 |
| 40 | +- Fixed issue with building package in tox environments |
| 41 | + |
| 42 | +## [0.2.11] - 2025-12-09 |
| 43 | + |
| 44 | +### Added |
| 45 | + |
| 46 | +- Added model metadata to output |
| 47 | +- Added more classes to `__init__.py` for easier imports |
| 48 | +- Added skipping of unprocessable inputs |
| 49 | + |
| 50 | +### Fixed |
| 51 | + |
| 52 | +- Loading multiple GPUs in parallel processes was not possible |
| 53 | +- Fixed hanging problem after error in processing occurred |
| 54 | + |
| 55 | +### Changed |
| 56 | + |
| 57 | +- Renamed many of the classes and functions for better clarity |
| 58 | + |
| 59 | +## [0.2.10] - 2025-11-28 |
| 60 | + |
| 61 | +### Added |
| 62 | + |
| 63 | +- Added support for Perch model v2 |
| 64 | + |
| 65 | +### Fixed |
| 66 | + |
| 67 | +- Removed support for LiteRT on macOS ARM64 due to incompatibility issues |
| 68 | + |
| 69 | +## [0.2.9] - 2025-11-27 |
| 70 | + |
| 71 | +### Added |
| 72 | + |
| 73 | +- Added option to supervise progress using callback function during inference |
| 74 | + |
| 75 | +## [0.2.8] - 2025-11-26 |
| 76 | + |
| 77 | +### Added |
| 78 | + |
| 79 | +- Added option to predict and encode raw audio numpy arrays |
| 80 | + |
| 81 | +## [0.2.7] - 2025-11-25 |
| 82 | + |
| 83 | +### Added |
| 84 | + |
| 85 | +- Added parameter `speed` to control playback speed of audio during inference |
| 86 | + |
| 87 | +## [0.2.5] - 2025-11-17 & [0.2.6] - 2025-11-21 |
| 88 | + |
| 89 | +### Bugfix |
| 90 | + |
| 91 | +- Fixed issue with using ProtoBuf CPU backend and TensorFlow GPU being available |
| 92 | +- Fixed #17: Issue on macOS with too long ring buffer names |
| 93 | +- Fixed #19: `queue.qsize()` is not used anymore |
| 94 | +- Fixed issue with hanging session because of logging |
| 95 | +- Fixed issue with downloading same model simultaneously |
| 96 | + |
| 97 | +### Changed |
| 98 | + |
| 99 | +- Removed `litert` install option, now litert is always installed if possible |
| 100 | +- Rename `tf` library to `tflite` to better reflect the usage of TFLite/LiteRT |
| 101 | +- Improved prediction speed, esp. for half-precision models (+10 seg/s) |
| 102 | +- Lowered dependencies |
| 103 | +- Update `ai-edge-litert` to version 2.0.3 on `repro` |
| 104 | +- Better download progress indication of model files |
| 105 | +- Model loading in tests is done before running other tests |
| 106 | + |
| 107 | +### Added |
| 108 | + |
| 109 | +- Added `repro` option to be able to get reproducible results |
| 110 | +- Added support for Python 3.13 |
| 111 | +- Added CI on GitHub Actions for testing on multiple OS and Python versions |
| 112 | + |
| 113 | +### Removed |
| 114 | + |
| 115 | +- Remove unused dependencies `numba` and `resampy` |
| 116 | + |
| 117 | +## [0.2.4] - 2025-11-05 |
| 118 | + |
| 119 | +### Bugfixes |
| 120 | + |
| 121 | +- Fixed issue with loading supported files from a folder |
| 122 | + |
| 123 | +### Changed |
| 124 | + |
| 125 | +- Increased half-precision prediction speed |
| 126 | +- Set "pyarrow==22.0.0" |
| 127 | +- Set "numpy==2.0.2" because of compatibility with `perch-hoplite` |
| 128 | +- Set default "half_precision" parameter to False because of lower speed |
| 129 | + |
| 130 | +### Added |
| 131 | + |
| 132 | +- Add half precision to CLI |
| 133 | + |
| 134 | +## [0.2.3] - 2025-11-04 |
| 135 | + |
| 136 | +### Added |
| 137 | + |
| 138 | +- Added support for Python 3.12 |
| 139 | + |
| 140 | +### Changed |
| 141 | + |
| 142 | +- Changed tensorflow to newest version 2.20.0 |
| 143 | + |
| 144 | +## [0.2.2] - 2025-11-03 |
| 145 | + |
| 146 | +### Added |
| 147 | + |
| 148 | +- Added support for running multiple sessions in a row or in parallel using threading or multiprocessing |
| 149 | +- Each session has its own logger and log file |
| 150 | + |
| 151 | +### Changed |
| 152 | + |
| 153 | +- Changed naming of the benchmark output files |
| 154 | + |
| 155 | +## [0.2.1] - 2025-10-29 |
| 156 | + |
| 157 | +### Added |
| 158 | + |
| 159 | +- Added parameter `is_raven` to load function to specify whether a custom Protobuf model is a Raven model or not |
| 160 | +- Fix int8 acoustic model wrong inference parameters |
| 161 | +- Added tests |
| 162 | + |
| 163 | +## [0.2.0] - 2025-10-27 |
| 164 | + |
| 165 | +### Changed |
| 166 | + |
| 167 | +- Refactored the whole codebase to be able to load model and predict scores in two separate steps |
| 168 | + |
| 169 | +## [0.2.0a0] - 2025-07-29 |
| 170 | + |
| 171 | +### Changed |
| 172 | + |
| 173 | +- Refactored the whole codebase |
| 174 | + |
| 175 | +## [0.1.7] - 2025-03-19 |
| 176 | + |
| 177 | +### Changed |
| 178 | + |
| 179 | +- Switched model download links from TUCcloud to Zenodo [#10](https://github.com/birdnet-team/birdnet/issues/10) |
| 180 | + |
| 181 | +### Fixed |
| 182 | + |
| 183 | +- Added check for mono files [#9](https://github.com/birdnet-team/birdnet/issues/9) |
| 184 | + |
| 185 | +## [0.1.6] - 2024-09-04 |
| 186 | + |
| 187 | +### Added |
| 188 | + |
| 189 | +- Support for multiprocessing using `predict_species_within_audio_files_mp` |
| 190 | + |
| 191 | +### Changed |
| 192 | + |
| 193 | +- Separate `ModelV2M4TFLite` into `AudioModelV2M4TFLite` and `MetaModelV2M4TFLite` |
| 194 | +- Separate `ModelV2M4Protobuf` into `AudioModelV2M4Protobuf` and `MetaModelV2M4Protobuf` |
| 195 | +- Separate `ModelV2M4` into `AudioModelV2M4` and `MetaModelV2M4` |
| 196 | +- Move v2.4 models to `birdnet.models.v2m4` |
| 197 | +- Yield results of `predict_species_within_audio_file` instead of returning an OrderedDict |
| 198 | +- Extracted method `predict_species_within_audio_file` and `predict_species_at_location_and_time` from their respective model |
| 199 | +- set default value for `batch_size` to 100 |
| 200 | + |
| 201 | +## [0.1.5] - 2024-08-16 |
| 202 | + |
| 203 | +### Fixed |
| 204 | + |
| 205 | +- Custom Raven audio model didn't return same results as custom TFLite model because of sigmoid layer |
| 206 | +- TFLite meta model was not returning correct results |
| 207 | + |
| 208 | +### Changed |
| 209 | + |
| 210 | +- Rename `CustomModelV2M4TFLite` to `CustomAudioModelV2M4TFLite` |
| 211 | +- Rename `CustomModelV2M4Raven` to `CustomAudioModelV2M4Raven` |
| 212 | + |
| 213 | +## [0.1.4] - 2024-08-13 |
| 214 | + |
| 215 | +### Added |
| 216 | + |
| 217 | +- Support to load custom TFLite models using `CustomModelV2M4TFLite` |
| 218 | +- Support to load custom Raven (Protobuf) models using `CustomModelV2M4Raven` |
| 219 | + |
| 220 | +## [0.1.3] - 2024-08-13 |
| 221 | + |
| 222 | +### Changed |
| 223 | + |
| 224 | +- Make CUDA dependency optional, install with `birdnet[and-cuda]` |
| 225 | + |
| 226 | +### Fixed |
| 227 | + |
| 228 | +- Bugfix 'ERROR: Could not find a version that satisfies the requirement nvidia-cuda-nvcc-cu12 (Mac/Ubuntu/Windows)' (#4) |
| 229 | + |
| 230 | +## [0.1.2] - 2024-08-07 |
| 231 | + |
| 232 | +### Added |
| 233 | + |
| 234 | +- Add GPU support by introducing the Protobuf model (v2.4) |
| 235 | + |
| 236 | +### Changed |
| 237 | + |
| 238 | +- Rename class 'ModelV2M4' to 'ModelV2M4TFLite' |
| 239 | +- 'ModelV2M4' defaults to Protobuf model now |
| 240 | +- Sorting of prediction scores is now: score (desc) & name (asc) |
| 241 | + |
| 242 | +### Fixed |
| 243 | + |
| 244 | +- Bugfix output interval durations are now always of type 'float' |
| 245 | + |
| 246 | +## [0.1.1] - 2024-08-02 |
| 247 | + |
| 248 | +### Added |
| 249 | + |
| 250 | +- Add parameter 'chunk_overlap_s' to define overlapping between chunks (#3) |
| 251 | + |
| 252 | +### Removed |
| 253 | + |
| 254 | +- Remove parameter 'file_splitting_duration_s' instead load files in 3s chunks (#2) |
| 255 | +- Remove 'librosa' dependency |
| 256 | + |
| 257 | +## [0.1.0] - 2024-07-23 |
| 258 | + |
| 259 | +- Initial release |
| 260 | + |
| 261 | +[Unreleased]: https://github.com/birdnet-team/birdnet/compare/v0.2.14...HEAD |
| 262 | +[0.2.14]: https://github.com/birdnet-team/birdnet/compare/v0.2.13...v0.2.14 |
| 263 | +[0.2.13]: https://github.com/birdnet-team/birdnet/compare/v0.2.12...v0.2.13 |
| 264 | +[0.2.12]: https://github.com/birdnet-team/birdnet/compare/v0.2.11...v0.2.12 |
| 265 | +[0.2.11]: https://github.com/birdnet-team/birdnet/compare/v0.2.10...v0.2.11 |
| 266 | +[0.2.10]: https://github.com/birdnet-team/birdnet/compare/v0.2.9...v0.2.10 |
| 267 | +[0.2.9]: https://github.com/birdnet-team/birdnet/compare/v0.2.8...v0.2.9 |
| 268 | +[0.2.8]: https://github.com/birdnet-team/birdnet/compare/v0.2.7...v0.2.8 |
| 269 | +[0.2.7]: https://github.com/birdnet-team/birdnet/compare/v0.2.6...v0.2.7 |
| 270 | +[0.2.6]: https://github.com/birdnet-team/birdnet/compare/v0.2.5...v0.2.6 |
| 271 | +[0.2.5]: https://github.com/birdnet-team/birdnet/compare/v0.2.4...v0.2.5 |
| 272 | +[0.2.4]: https://github.com/birdnet-team/birdnet/compare/v0.2.3...v0.2.4 |
| 273 | +[0.2.3]: https://github.com/birdnet-team/birdnet/compare/v0.2.2...v0.2.3 |
| 274 | +[0.2.2]: https://github.com/birdnet-team/birdnet/compare/v0.2.1...v0.2.2 |
| 275 | +[0.2.1]: https://github.com/birdnet-team/birdnet/compare/v0.2.0...v0.2.1 |
| 276 | +[0.2.0]: https://github.com/birdnet-team/birdnet/compare/v0.2.0a0...v0.2.0 |
| 277 | +[0.2.0a0]: https://github.com/birdnet-team/birdnet/compare/v0.1.7...v0.2.0a0 |
| 278 | +[0.1.7]: https://github.com/birdnet-team/birdnet/compare/v0.1.6...v0.1.7 |
| 279 | +[0.1.6]: https://github.com/birdnet-team/birdnet/compare/v0.1.5...v0.1.6 |
| 280 | +[0.1.5]: https://github.com/birdnet-team/birdnet/compare/v0.1.4...v0.1.5 |
| 281 | +[0.1.4]: https://github.com/birdnet-team/birdnet/compare/v0.1.3...v0.1.4 |
| 282 | +[0.1.3]: https://github.com/birdnet-team/birdnet/compare/v0.1.2...v0.1.3 |
| 283 | +[0.1.2]: https://github.com/birdnet-team/birdnet/compare/v0.1.1...v0.1.2 |
| 284 | +[0.1.1]: https://github.com/birdnet-team/birdnet/compare/v0.1.0...v0.1.1 |
| 285 | +[0.1.0]: https://github.com/birdnet-team/birdnet/releases/tag/v0.1.0 |
0 commit comments