/S2L_45

MIT LicenseMIT

SOUNDSCAPES TO LANDSCAPES (S2L) DATASET

S2L_45 Dataset (06/06/2021)

The S2L_45 dataset was created as a part of a larger effort of monitoring biodiversity through the Soundscapes2Landscapes (S2L) project. The project seeks to advance animal biodiversity monitoring by making use of Earth-observing satellites. The availability of inexpensive MEMS sound recorders that capture audio of sufficient quality has made possible the scaling of research from single locations and organism, to full animal communities across landscapes. Indeed, the broad geographic scale of the Soundscapes to Landscapes (S2L) project is pioneering.

The dataset consists of two parts:

  1. ROI set - Regions of Interests with labels are annotated by expert birders and bounded in time and frequency.
  2. Golden Validation (GV) set - One-minute clips with minute-level coarse labeling of all the species present.
  • Number of Classes: 45
  • Total Number of ROIs: 34,968
  • Total Number of GV-set clips: 901
  • Location: Sonoma County, CA, USA
  • Sites: 586
  • Audio Recorder: AudioMoth
  • Sample Frequency: 48,000 Hz
Sr.No. CLASS ROIs GV_instances min_dur (sec) max_dur (sec) average_dur (sec) avg_min_freq (Hz) avg_max_freq (Hz)
1 ACWO 528 67 0.09 0.48 0.22 945.43 3820.5
2 AMCR 553 61 0.28 0.53 0.39 633.96 2607.36
3 AMRO 1364 53 0.19 3.5 0.85 1331.11 7390.21
4 ATFL 445 2 0.45 0.78 0.53 1376.67 3603.96
5 BEWR 503 87 0.47 3.88 0.67 3464.94 6181.01
6 BGGN 966 12 0.12 0.86 0.63 2053.58 11104.6
7 BHGR 866 38 0.12 0.69 0.6 1340.7 2842.83
8 BRCR 609 42 1.28 1.97 1.91 4036.42 8043.72
9 BTPI 561 6 0.59 1.1 0.72 54.63 589.51
10 BTYW 699 16 1.5 2.84 1.64 2910.01 8054.39
11 CAQU 1083 30 0.22 0.48 0.32 1208.05 2155.49
12 CASJ 604 44 0.38 1.75 1.15 976.3 4887.09
13 CAVI 642 23 0.28 0.66 0.37 2322.59 4989.02
14 CBCH 607 67 0.22 1.27 0.39 5449.34 7476.58
15 CORA 536 47 0.72 4.43 1.06 347.69 1949.83
16 COYE 1038 13 1.59 2.28 1.98 1893.88 7409.39
17 DEJU 1345 85 0.39 2.66 2.12 2483.54 6705.8
18 GHOW 474 3 2.81 3.1 2.87 0.0 647.17
19 HOFI 571 35 0.19 0.25 0.22 2920.29 4397.6
20 LAZB 1530 4 2.11 2.16 2.13 2250.0 8064.0
21 MAWR 515 6 1.72 2.28 1.87 3486.1 9626.83
22 MODO 552 27 2.31 3.58 3.1 190.1 934.37
23 MOUQ 520 5 0.41 0.45 0.43 562.5 1969.0
24 NOFL 824 20 2.09 7.02 3.3 1981.23 3180.94
25 NUWO 562 11 0.31 3.95 0.39 2638.64 4359.42
26 OATI 518 65 0.31 0.39 0.35 4720.99 7149.5
27 OCWA 1334 89 1.19 3.31 1.94 2959.76 6618.77
28 PAWR 695 17 5.22 5.34 5.25 4613.06 8451.05
29 PIWO 585 5 0.86 4.14 1.49 1392.87 2341.35
30 PSFL 2993 125 0.34 0.69 0.38 3875.98 7237.3
31 RSHA 919 3 0.38 2.06 0.7 866.0 3437.37
32 RWBL 891 69 0.84 2.58 0.89 2165.66 3666.09
33 SAVS 563 13 1.59 1.62 1.61 4124.0 10216.0
34 SOSP 1144 39 0.44 3.43 0.56 1372.55 7328.77
35 SPTO 819 88 0.6 1.31 1.13 2992.58 7287.75
36 STJA 694 90 0.3 1.75 1.6 1576.9 4936.31
37 SWTH 605 8 2.41 2.44 2.42 1312.0 6376.0
38 WAVI 607 53 0.38 4.5 2.1 2575.16 6387.19
39 WBNU 513 21 0.12 0.53 0.44 1458.86 4009.55
40 WCSP 500 25 2.48 2.53 2.5 2344.0 6656.0
41 WEME 964 41 0.31 3.01 0.67 1379.96 3327.64
42 WETA 532 12 0.22 0.77 0.32 2077.6 4207.02
43 WITU 471 18 0.34 2.09 1.45 240.07 3086.62
44 WIWA 507 40 1.19 2.28 1.41 3915.6 8098.56
45 WREN 617 30 2.09 2.92 2.37 1835.12 3580.26
Total: 34698 1655

Classnames are Four-letter (English Name) Alpha Codes for Bird Species, in accordance with the 61st AOU Supplement (2020). More details: The Institute for Bird Populations and Species List.

Download

The dataset will be made public soon.

Results

Baseline:

Licence

MIT Licence