Published February 16, 2024 | Version v1
Dataset Open

Caltech Fish Counting - Domain Adaptive Object Detection Dataset 2024

Kay, Justin1
Haucke, Timm1
Stathatos, Suzanne
Deng, Siqi2
Young, Erik3
Perona, Pietro4 ORCID icon
Beery, Sara1
Van Horn, Grant5
  • 1. ROR icon Massachusetts Institute of Technology
  • 2. ROR icon Amazon (United States)
  • 3. Skagit Fisheries Enhancement Group
  • 4. ROR icon California Institute of Technology
  • 5. ROR icon University of Massachusetts Amherst

Description

A full instructional guide is provided here

CFC-DAOD is an expansion of the Caltech Fish Counting Dataset for detecting, tracking, and counting fish in sonar videos. This dataset is specifically focused on domain-adaptive object detection.

Files

cfc_train.zip
Files (27.5 GB)
Name Size
cfc_train.zip
md5:935b4cd5ae5812035051f24e6707ee17
16.9 GB Preview Download
cfc_val.json
md5:0e559676de668a0bbc8391323d8b450f
11.4 MB Preview Download
cfc_val.zip
md5:e662ae8318621d1a636f0befadddaf48
4.4 GB Preview Download
cfc_channel_train.zip
md5:d17e0485674327df3d7611a5d6b999b1
3.1 GB Preview Download
cfc_channel_test.zip
md5:9c15b9c9dc6784cce9dba21e81cb514a
3.0 GB Preview Download
cfc_train.json
md5:57ce891a14b8aa71699988ccd488ccf4
53.1 MB Preview Download
cfc_channel_train.json
md5:c194923784eb8e5d64873a16c543ee64
17.2 MB Preview Download
cfc_channel_test.json
md5:9b701b4b884c4061425dd82450c9974b
14.4 MB Preview Download

Additional details

Additional titles Funding Dates
Created:
February 16, 2024
Modified:
July 10, 2024