Birdvoxdetect

This is the fourth alpha release of the 0.5 series.
This release fixes a bug: "cannot unpack non-iterable NoneType object" (82), which arose when calling BirdVoxClassify on empty PCEN frames (PR 83).

This is the third alpha release of the 0.5 series.
This release fixes a bug introduced by 73, which caused a `FutureWarning` in Pandas: `Sorting because non-concatenation axis is not aligned` (commit 7ba070d).

This is the second alpha release of the 0.5 series.
This release fixes a bug introduced by 73, which caused missing columns in the checklist when no flight calls were present in the first 30 minutes (79).

This is the first alpha release of the 0.5 series

The main change with respect to the v0.4 series is a modification of the flight call classification algorithm which enforces hierarchical consistency in the TaxoNet output. We now ensure that the predicted species belongs to the predicted family and that the predicted family belongs to the predicted order.
See https://github.com/BirdVox/birdvoxdetect/pull/73 for a description of the algorithm

Additionally, BirdVoxDetect now reports order confidence, family confidence, and species confidence in addition to detection confidence.

This is the first bugfix release of the 0.4 series.
Fixed:
* 76 KeyError: 'Species (4-letter code)' when printing checklist count of an empty file

This is the first stable release of the 0.4 series. It is functionally identical to 0.4.0b2.

The main change with respect to the v0.3 series is the improved species classifier (TaxoNet). Furthermore, each detected event now contains predictions at coarser taxonomical levels (family and order) in addition to species.

The process_file function now supports a new keyword argument, predict_proba, defaulting to False. If set to true, BirdVoxDetect will export a JSON file containing probabilistic predictions for every species, family, and order in TaxoNet. The JSON file also contains a copy of the BirdVoxClassify taxonomy, a copy of the sensor fault log given by BIrdVoxActivate, and additional metadata: file name, audio duration, MD5 checksum, package versions, hardware specifications, and elapsed time.

We have provided a command-line interface for the predict-proba keyword argument, via the optional flag --predict-proba.