Birdvoxdetect

This is the first alpha release of the 0.6 series.

The main change with respect to the v0.5 series is the update from BirdVoxClassify v0.2 to v0.3. The species classifier is now "flat multitask" by default and includes hierarchical consistency. Furthermore, the order "Passeriforme" is now spelled in the plural: "Passeriformes". We have added various encodings for species: English name, scientific (Latin) name, and 4-letter code (85). The "other" class is represented as an empty string. Lastly, we have truncated the number of digits of confidence in the checklist (86).

This is the first bugfix release of the 0.5 series.

In comparison with 0.5.0, we have changed the default classifier: it is no longer TaxoNet but "flat multitask". In doing so, we revert to the species classifier that was already present in BirdVoxDetect 0.3.0. We have made this change because we have noticed a surprisingly poor performance of TaxoNet on hold-out data. Thus, we err on the side of caution and offer a more straightforward deep learning model as the new default. We are currently investigating the cause of the problem in coordination with BirdVoxClassify developers.

This is the first stable 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. Additionally, BirdVoxDetect now reports order confidence, family confidence, and species confidence in addition to detection confidence. More details below.

Our numerical experiments on 300 hours of annotated audio have demonstrated that TaxoNet outperforms the classifier of BirdVoxDetect v0.3 (see previous release notes for details). However, we have noted that the output of TaxoNet is occasionally inconsistent across levels. For example, the "family" level may return "Turdidae" (Thrushes) while the "species" level would return "WTSP", i.e. White-Throated Sparrow.

Although infrequent, this lack of consistency may be confusing to users. To address it, we have implemented a new algorithm for hierarchical classification which ensures consistency by design: see https://github.com/BirdVox/birdvoxdetect/pull/73

The key idea is to query TaxoNet in a coarse-to-fine fashion, while progressively narrowing down the list of candidate species corresponding to the detected flight call. Thanks to Justin Salamon for proposing this elegant solution.

We have taken the opportunity of rewriting the classify_species function (see 73) to enrich the information that is stored in the BirdVoxDetect checklist. In addition to textual labels for order, family, and species, we have added three new columns, corresponding to "order confidence", "family confidence" , and "species confidence".

These values of confidence all range between 50% and 100% and provide insight on the prediction of TaxoNet. Users should be warned, however, that the percentages are not calibrated, and thus do not come with any probabilistic interpretation. Rather, they are simply the output values of the sigmoid units in TaxoNet. This is unlike "Detection confidence (%)", which is carefully calibrated to match the precision of the detector (on a precision-recall curve).

This is the second beta release of the 0.5 series. It is functionally identical to v0.5.0b1, with small fixes to the README page.

This is the first beta 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 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 fifth alpha release of the 0.5 series.
This release fixes a bug which arose when calling BirdVoxClassify on empty PCEN frames (PR 84, sequel of 83).