Account Creation Issues: We have received reports of issues with creating new user accounts and linking accounts to CAM, and are currently investigating the root cause. In the meantime: - If you're experiencing errors creating new users, please use the "Quick Add" feature instead (click the "Quick Add" button on the Manage Users page). - If you're experiencing errors linking CAM accoun...
View All
Search Help Show/Hide Menu
Completion Rubric
64%
22.25 / 35
View Report
Alaska Fisheries Science Center Metadata Librar... > Resource Assessment and Conservation Engineerin... > Midwater Assessment and Conservation Engineerin... > Conservation Engineering Research >

AFSC/RACE/MACE: 2019-2020 Salmon and Pollock Object Detection

Data Set (DS)  |  Alaska Fisheries Science Center (AFSC)
GUID: gov.noaa.nmfs.inport:69412  |  Updated: January 30, 2026  |  Published / External 
Summary
Item Identification
Keywords
Physical Location
Data Set Info
Support Roles
Extents
Access Info
Distribution Info
URLs
Data Quality
Data Management
Lineage
Acquisition Info
Child Items
Related Items
Catalog Details

Summary

Short Citation
Alaska Fisheries Science Center, 2026: AFSC/RACE/MACE: 2019-2020 Salmon and Pollock Object Detection, https://www.fisheries.noaa.gov/inport/item/69412.
Full Citation Examples

Abstract

This dataset contains the annotated data, object detection models, and video data used to fine-tune and evaluate object detection models for salmon and pollock detection in videos collected in a commercial pollock trawl in Alaska. The annotated data contains track annotations for all Pacific salmon (Oncorhynchus spp.) and walleye pollock (Gadus chalcogrammus) present and frame level annotations that describe the trawl and background conditions. The annotated data is composed of 11,572 salmon and 73,394 pollock annotations from 16,989 video frames that came from 184 video clips. Five-fold cross validation was used to evaluate EfficientDet D2 and YOLO11n object detection models, and this dataset includes all of these models and a single-class, salmon-only YOLO11n model. In addition to the annotated video frames that are provided as images, videos from three fishing tows that were not annotated are included with their respective YOLO11n detections. These three fishing tows were used to further evaluate the object detection performance of the best performing pollock and salmon model and the salmon-only model.

Purpose

Bycatch reduction devices (BRDs) are used in the Alaska walleye pollock (Gadus chalcogrammus) fishery to reduce Pacific salmon (Oncorhynchus spp.) bycatch. Evaluation of BRD effectiveness often requires people to process collected or live-feed video, which can be a tedious, expensive, and time-consuming task. Deep learning can be used to automate the detection and classification of fish in videos to support BRD and other fisheries bycatch work. This dataset was developed to evaluate the performance of widely-available, open-source pre-trained object detection models for the task of pollock and salmon detection.

Data Access & Downloads

  • 2019-2020 Salmon and Pollock Object Detection dataset
    PNG - Portable Network Graphics Format, <2MB

    This folder within the Alaska Fisheries Science Center's Google cloud storage contains the data and object detection models used for the Wilson et al. 2025 study on automated salmon and pollock detection in trawl videos from the Alaska commercial pollock fishery.

    The 'annotated_data' folder contains a folder for each of the 184 video clips used in the study that provides video frames as png files and the annotations as a custom COCO json file.

    The 'tow_data' folder contains videos as .mov files and csv files of fish detections for three fishing tows that were also used in the study to further evaluate object detection model performance.

    The 'models' folder contains the object detection models that were used in the study.

    The 'documentation' folder contains text files that provide details about the data and the cloud storage file structure.

Access Constraints:

If using these data, please cite https://doi.org/10.1093/icesjms/fsaf168.

URLs

Child Items

Type Title
Data Set Salmon tracking

Contact Information

Point of Contact
Katherine Wilson
katherine.wilson@noaa.gov
2065264474

Extents

Geographic Area 1

-166.03722222222° W, -165.59805555556° E, 54.69833333333° N, 54.32472222222° S

Time Frame 1
2020-08-07 - 2020-08-10

Item Identification

Title: AFSC/RACE/MACE: 2019-2020 Salmon and Pollock Object Detection
Short Name: 2019-2020 Salmon Detection
Creation Date: 2023-03-08
Revision Date: 2025-09-25
Abstract:

This dataset contains the annotated data, object detection models, and video data used to fine-tune and evaluate object detection models for salmon and pollock detection in videos collected in a commercial pollock trawl in Alaska. The annotated data contains track annotations for all Pacific salmon (Oncorhynchus spp.) and walleye pollock (Gadus chalcogrammus) present and frame level annotations that describe the trawl and background conditions. The annotated data is composed of 11,572 salmon and 73,394 pollock annotations from 16,989 video frames that came from 184 video clips. Five-fold cross validation was used to evaluate EfficientDet D2 and YOLO11n object detection models, and this dataset includes all of these models and a single-class, salmon-only YOLO11n model. In addition to the annotated video frames that are provided as images, videos from three fishing tows that were not annotated are included with their respective YOLO11n detections. These three fishing tows were used to further evaluate the object detection performance of the best performing pollock and salmon model and the salmon-only model.
Purpose:

Bycatch reduction devices (BRDs) are used in the Alaska walleye pollock (Gadus chalcogrammus) fishery to reduce Pacific salmon (Oncorhynchus spp.) bycatch. Evaluation of BRD effectiveness often requires people to process collected or live-feed video, which can be a tedious, expensive, and time-consuming task. Deep learning can be used to automate the detection and classification of fish in videos to support BRD and other fisheries bycatch work. This dataset was developed to evaluate the performance of widely-available, open-source pre-trained object detection models for the task of pollock and salmon detection.
Notes:

The Wilson et al. 2025 study found that the YOLO models performed better than EfficientDet and on average detected 90% of salmon and pollock with 72% accuracy using a 50% detection overlap threshold.

Keywords

Theme Keywords

Thesaurus Keyword
UNCONTROLLED
None detection
None pollock
None salmon
None tracking
None trawls
None video

Physical Location

Organization: Alaska Fisheries Science Center
City: Seattle
State/Province: WA
Country: USA
Location Description:

RACE division, MACE program

Data Set Information

Data Set Scope Code: Data Set
Data Set Type: Imagery
Maintenance Frequency: As Needed
Data Presentation Form: Video (digital)
Entity Attribute Overview:

The dataset includes the subset of video clips that were collected in the commercial pollock fishery in Alaska and annotated for salmon and pollock, the object detection models that were trained and evaluated, all the video recorded for 3 fishing tows that were used to further evaluate model performance, and the model detections for these 3 fishing tows.
Data Set Credit: F/V Pacific Explorer, David Bryan, Deborah Sharpe , Katherine Hellen-Schneider, Matthew Callahan, Connor Fox

Support Roles

Author

CC ID: 1222689
Date Effective From: 2023
Date Effective To:
Contact (Person): Wilson, Katherine
Address: 7600 Sand Point Way NE, Bldg. 4
Seattle, FL 98115
United States
Email Address: katherine.wilson@noaa.gov
Phone: 2065264474

Co-Author

CC ID: 1222687
Date Effective From: 2020
Date Effective To:
Contact (Organization): Pacific States Marine Fisheries Commission (PSMFC)
Address:
Contact Instructions:

Lurbur, Moses <mlurbur@gmail.com>

Co-Author

CC ID: 1222688
Date Effective From: 2019
Date Effective To:
Contact (Person): Yochum, Noelle
Address: 7600 Sand Point Way NE, Bldg. 4
Seattle, WA 98115
United States
Email Address: noelle.yochum@noaa.gov
Phone: (206) 526-4432
Mobile: (916) 719-5920

Point of Contact

CC ID: 1463686
Date Effective From: 2025
Date Effective To:
Contact (Person): Wilson, Katherine
Address: 7600 Sand Point Way NE, Bldg. 4
Seattle, FL 98115
United States
Email Address: katherine.wilson@noaa.gov
Phone: 2065264474

Extents

Extent Group 1

Extent Description:

All 7 tows from Trip 1 were within this extent.

Extent Group 1 / Geographic Area 1

CC ID: 1222708
W° Bound: -166.03722222222
E° Bound: -165.59805555556
N° Bound: 54.69833333333
S° Bound: 54.32472222222

Extent Group 1 / Time Frame 1

CC ID: 1222709
Time Frame Type: Range
Start: 2020-08-07
End: 2020-08-10

Extent Group 2

Extent Description:

All 4 tows from trip 2 were within these bounds.

Extent Group 2 / Geographic Area 1

CC ID: 1222710
W° Bound: -165.72888888889
E° Bound: -165.43055555556
N° Bound: 54.78472222222
S° Bound: 54.51527777778

Extent Group 2 / Time Frame 1

CC ID: 1222711
Time Frame Type: Range
Start: 2020-08-13
End: 2020-08-16

Extent Group 3

Extent Description:

All 26 tows were within this extent.

Extent Group 3 / Geographic Area 1

CC ID: 1222713
W° Bound: -166
E° Bound: -164.5
N° Bound: 55.33333333333
S° Bound: 54.33333333333

Extent Group 3 / Time Frame 1

CC ID: 1222714
Time Frame Type: Range
Start: 2019-06-01
End: 2019-06-30

Access Information

Data License: CC0-1.0
Data License URL: https://creativecommons.org/publicdomain/zero/1.0/
Data License Statement: These data were produced by NOAA and are not subject to copyright protection in the United States. NOAA waives any potential copyright and related rights in these data worldwide through the Creative Commons Zero 1.0 Universal Public Domain Dedication (CC0-1.0).
Security Class: Unclassified
Data Access Procedure:

Video clip frames, annotations, object detection models, full tow videos and their detections can be accessed at https://console.cloud.google.com/storage/browser/nmfs_odp_afsc/RACE/MACE/salmon_pollock_object_detection.
Data Access Constraints:

If using these data, please cite https://doi.org/10.1093/icesjms/fsaf168.

Distribution Information

Distribution 1

CC ID: 1449416
Download URL: https://console.cloud.google.com/storage/browser/nmfs_odp_afsc/RACE/MACE/salmon_pollock_object_detection/
Distributor:
File Name: 2019-2020 Salmon and Pollock Object Detection dataset
Description:

This folder within the Alaska Fisheries Science Center's Google cloud storage contains the data and object detection models used for the Wilson et al. 2025 study on automated salmon and pollock detection in trawl videos from the Alaska commercial pollock fishery.

The 'annotated_data' folder contains a folder for each of the 184 video clips used in the study that provides video frames as png files and the annotations as a custom COCO json file.

The 'tow_data' folder contains videos as .mov files and csv files of fish detections for three fishing tows that were also used in the study to further evaluate object detection model performance.

The 'models' folder contains the object detection models that were used in the study.

The 'documentation' folder contains text files that provide details about the data and the cloud storage file structure.
Distribution Format: PNG - Portable Network Graphics Format
File Size: <2MB
Compression: Uncompressed

URLs

URL 1

CC ID: 1427611
URL: https://console.cloud.google.com/storage/browser/nmfs_odp_afsc/RACE/MACE/salmon_pollock_object_detection
Name: Dataset
Description:

Dataset that includes annotated data, models, and tow videos and detections used for the Wilson et al. 2025 study.

URL 2

CC ID: 1427612
URL: https://doi.org/10.1093/icesjms/fsaf168
Name: Automated fish detection in videos to support commercial fishing sustainability and innovation in the Alaska walleye pollock (Gadus chalcogrammus) trawl fishery
Description:

Scientific publication

URL 3

CC ID: 1449410
URL: https://github.com/noaa-afsc-mace/salmon_presence
Name: Salmon presence repo
Description:

Source code for salmon presence prediction and evaluation.

Data Quality

Representativeness:

Video from within the trawl of the commercial pollock fishery in Alaska and the associated annotations of salmon and pollock.
Accuracy:

See Wilson et. 2025 publications for details.
Analytical Accuracy:

See Wilson et. 2025 publications for details.
Horizontal Positional Accuracy:

See Wilson et. 2025 publications for details.
Vertical Positional Accuracy:

See Wilson et. 2025 publications for details.
Quantitation Limits:

See Wilson et. 2025 publications for details.
Bias:

See Wilson et. 2025 publications for details.
Comparability:

See Wilson et. 2025 publications for details.
Completeness Measure:

See Wilson et. 2025 publications for details.
Precision:

See Wilson et. 2025 publications for details.
Analytical Precision:

See Wilson et. 2025 publications for details.
Field Precision:

See Wilson et. 2025 publications for details.
Sensitivity:

See Wilson et. 2025 publications for details.
Detection Limit:

See Wilson et. 2025 publications for details.
Completeness Report:

See Wilson et. 2025 publications for details.
Conceptual Consistency:

See Wilson et. 2025 publications for details.
Quality Control Procedures Employed:

Annotations were done by trained personnel and reviewed by a lead scientist.

Data Management

Have Resources for Management of these Data Been Identified?: Yes
Approximate Percentage of Budget for these Data Devoted to Data Management: 0
Do these Data Comply with the Data Access Directive?: No
Is Access to the Data Limited Based on an Approved Waiver?: No
Actual or Planned Long-Term Data Archive Location: Other

Lineage

Lineage Statement:

The video portion of this project was collected onboard the FV Pacific Explorer by Noelle Yochum and Katherine Hellen-Schneider in 2019 and by Connor Fox in 2020 with Sexton housings equipped with Mobius cameras. Processing of video (identifying salmon and entering data) was conducted by David Bryan, Michelle Dyroy, Noelle Yochum, and Katherine Wilson. Video annotation (bounding box tracks of salmon and pollock) was done by Moses Lurbur, Deborah Sharpe, Katherine Hellen-Schneider, and Matthew Callahan. Annotation QA/QC was conducted by Katherine Wilson.

Acquisition Information

Instruments

Instrument 1

CC ID: 1222715
Identifier: Sexton Video Camera System
Instrument / Gear: Instrument
Instrument Type: Video Recorder
Description:

The Sexton video camera system consists of an underwater housing rated to 1000m, 2 LED lights, a Mobius2 Action Camera, and a power supply (either 2 NiMH 19.2v or 2 Li-Ion 14.4v batteries). The Mobius2 camera can record 1800 HD video at 30 frames per second and is suitable for low light scenarios.

Child Items

Rubric scores updated every 15m

Rubric Score Type Title
64
Data Set Salmon tracking

Related Items

Item Type Relationship Type Title
Data Set (DS) Larger Work Citation 2019 Salmon Excluder Charter
Data Set (DS) Larger Work Citation 2020 Salmon Excluder Charter

Catalog Details

Catalog Item ID: 69412
GUID: gov.noaa.nmfs.inport:69412
Metadata Record Created By: Katherine Wilson
Metadata Record Created: 2023-03-09 02:35+0000
Metadata Record Last Modified By: Katherine Wilson
Metadata Record Last Modified: 2026-01-30 19:42+0000
Metadata Record Published: 2025-09-26
Owner Org: AFSC
Metadata Publication Status: Published Externally
Do Not Publish?: N
Metadata Last Review Date: 2025-09-09
Metadata Review Frequency: 1 Year
Metadata Next Review Date: 2026-09-09