Quantitative results for instance segmentation considering different ratio values, in terms of mIoU on the Youtube-VIS 2019 dataset.

Abstract DSMZ-MBNMS Outcrop Vent and Non Vent community composition from still images collected with ROV Hercules on NA117 contains the observations of community to the family level in 8 vent and non-vent zones around the Octopus Garden and Octocone. Also included are location, subtrate, zone type, TASSE variable values: Easterness, Mean Depth, Northerness, Topographic position index (RDMV), Slope, Standard Deviation, and Surface Depth, BRESS geomorphon, and averate temperature. Methods Observational data was collected by manual identification of 1-minute interval still images in the DSMZ-MBNSM NA117 videos available at: https://www.youtube.com/playlist?list=PL41b0O3MiKnZjVYKiC8zunwc7tGS9L2Az . Observations were identified with the aid of deep sea guides*: HURL, NOAA, and MBARI. BRESS and TASSE variables were extraction for bathymetric grids availble via Marine Geoscience Data System under Compilations as OctopusGarden_MBARI. Substrate was identifed manually. Temperature data was collected in NA-117 with ROV Hercules. Vent and Non-Vent zones were defined by temperature. Environmental and Community data for each image sample were complined in QGIS. Data is identifed as Q for qualitative and C for catagorical. Substrate Catagories: 0: Sheet, 1 Bare Rock, 2 Carbonate, 3 Blocky sheet pillow lobate, 4 Lobate, 5 Pillow, 6 Sediment, 7 Lobate Pillow Mix, 8 Lobate Broken Pillow Mix, 9 Lobate Pillow Contact, 10 Sediment Lobate Contact, 11 Lobate Sheet Mix, 12 Lobate Sediment Mix, 13 Lobate Sheet Contact, 14 Sediment Pillow Mix. BRESS Categories: 1 Flat, 2 Footslope, 3 Ridge, 4 Shoulder, 5 Slope, 6 Valley. Vent/Non Vent Catagories: 1 Non-Vent, 2 Vent. University of Hawai‘i Undersea Research Laboratory (HURL) Deep-sea Animal Identification Guide. Available from http://www.soest.hawaii.edu/ HURL/HURLarchive/guide.php US Department of Commerce, N. (2018, August 23). Data: Animal Identification Guide: NOAA Office of Ocean Exploration and Research. Retrieved October 26, 2020, from https://oceanexplorer.noaa.gov/okeanos/animal_guide/animal_guide.html Jacobsen Stout, N., L. Kuhnz, L. Lundsten, B. Schlining, K. Schlining, and S. von Thun (eds.). The Deep-Sea Guide (DSG). Monterey Bay Aquarium Research Institute (MBARI). Consulted on: 2022-07-01.

Mean and standard deviation of classifier accuracy across repeated training instances using each the three machine learning methods (compound index, pretrained CNN and trained CNN) at six different tasks. Accuracy is the proportion of one-minute recordings from the test data that were correctly classified. Methods where accuracy was reported as significantly higher by the ANOVA test are indicated in superscript next to the mean value for the respective method (A = highest group, B = second highest group, no letter = lowest group). The Random baseline accuracy indicates the expected accuracy of a model that performs random classification. N = 100 for all tasks, except the Fish diversity (Australia) and Depth (French Polynesia) tasks, where N = 32.