You are here

HiFi Stereo Bird Cocktail Party Challenge

banniere

Motivation

The goal of this innovative task is to evaluate to what extent the use of stereophony can improve the recognition performance of bird songs in soundscapes.

Test set

The test set (SBCP-TEST) is composed of 527 * 20 sec chorus files (~ 30 min), in STEREO 192 kHz SR 16 bits. It has been recorded during the lockdown spring 2020 into Domaine du Rayol, Côte d'Azur, France, by H. Glotin using super canon microphones and the Scarlet sound card. Microphone aperture is 17 cm, 120°.

Training set

The training set is at your choice (except a hand labelled version of SBCP-TEST dataset). You can use the training set of BirdCLEF 2021 main task as well as any other training set from previous LifeCLEF editions. You are also allowed to process self-supervised training, source separation or BSS on the SBCP-TEST dataset itself.

We suggest the challengers to use the ICML2103 ML4B training data set (H. Glotin et J. Sueur 2013, available here): 30 sec of samples for each of these 35 species, total duration 18min, monophone 16bits, 44.1kHz SR, extracted from F. Deroussen naturophonia.fr (Deroussen, 2001. Oiseaux des jardins de France. Nashvert Prod. ; Deroussen et Jiguet, 2006. La sonotheque du Museum: Oiseaux de France, les passereaux. Nashvert prod.). This training data includes the phylogenetic tree (Thuillier et al., Nature, 2011) of these 35 species to maybe regulate the Chorus analysis.

Another interesting training set could be the NIPS4Bplus training data set (Morfi, V., Bas, Y., Pamuła, H., Glotin, H., & Stowell, D. (2019). NIPS4Bplus: a richly annotated birdsong audio dataset. PeerJ Computer Science, 5, e223). NIPS4Bplus is the richly annotated birdsong audio dataset, that is comprised of recordings containing bird vocalisations along with their active species tags plus the temporal annotations acquired for them. This training set (687 + 1000 files, nearly 3 hours) has been designed for task such as training models for bird population monitoring, species classification, birdsong vocalisation detection and classification. It was collected in seven regions in France and Spain.

Task description

You have to predict in each 5 second chunk of each test file, the presence or absence of each of the 35 species listed below. All the run must have been fully automatic, and submitted with depicted script to be rerun for validation of the best run.

List of the 35 species to predict (in alphabetic order) for each chunck of 5 sec. is :
1 aegithalos_caudatus
2 alauda_arvensis
3 anthus_trivialis
4 branta_canadensis
5 carduelis_chloris
6 certhia_brachydactyla
7 columba_palumbus
8 corvus_corone
9 cuculus_canorus
10 dendrocopos_major
11 emberiza_citrinella
12 erithacus_rubecula
13 fringilla_coelebs
14 garrulus_glandarius
15 luscinia_megarhynchos
16 motacilla_alba
17 oriolus_oriolus
18 parus_caeruleus
19 parus_major
20 parus_palustris
21 pavo_cristatus
22 phasianus_colchicus
23 phoenicurus_phoenicurus
24 phylloscopus_collybita
25 picus_viridis
26 prunella_modularis
27 sitta_europaea
28 streptopelia_decaocto
29 strix_aluco
30 sturnus_vulgaris
31 sylvia_atricapilla
32 troglodytes_troglodytes
33 turdus_merula
34 turdus_philomelos
35 turdus_viscivorus

How to participate ?

1. Subscribe to LifeClef, Task 2 - BirdCLEF, by filling this form
2. Download SBCP-TEST dataset
3. Email your runs to glotin@univ-tln.fr in .csv format.
Each raw per file. One species per column x per chunck 0-5, 5-10, 10-15, 15-20
Each cell = probability that the species is present into the chunk.
You will also submit a decision file with the thresholded probability, same format, but 0 for absence, 1 for presence.

Exampe of a probability run :
HDBirdCocktailParty-001.wav, p(S1|C1),p(S1|C2),..., p(S1|C4), p(S2|C1),..., p(S35|C4)
...
HDBirdCocktailParty-527.wav, p(S1|C1),p(S1|C2),..., p(S1|C4), p(S2|C1),..., p(S35|C4)

Schedule

You can submit up to 5 runs by the 7th of may 2022 to glotin@univ-tln.fr
Possible corrections up to the 15th of may.

Credits

Challenge created by H. Glotin with Coraline Mounier & Noreen Blaukat, CNRS LIS DYNI Toulon.

Contact

glotin@univ-tln.fr