Purdue University

Edit Project

MSE Walls

Instance Segmentation

MSE Walls Image Dataset

Try Pre-Trained Model
Versions
2024-03-28 10:47pm
v8
Mar 29, 2024
2024-02-29 2:50pm
v7
Feb 29, 2024
2024-01-21 12:06pm
v6
Jan 21, 2024
2024-01-17 3:28pm
v5
Jan 17, 2024
2024-01-11 10:39pm
v4
Jan 12, 2024
2023-12-06 5:50pm
v3
Dec 6, 2023
2023-11-27 10:57pm
v2
Nov 28, 2023
2023-11-08 12:09am
v1
Nov 8, 2023
v8

2024-03-28 10:47pm

Generated on Mar 29, 2024
Download Dataset

6052 Total Images

View All Images
Annotation VisualizationDataset Image Annotation VisualizationDataset Image Annotation VisualizationDataset Image Annotation VisualizationDataset Image Annotation VisualizationDataset Image Annotation VisualizationDataset Image Annotation VisualizationDataset Image Annotation VisualizationDataset Image Annotation VisualizationDataset Image Annotation VisualizationDataset Image Annotation VisualizationDataset Image Annotation VisualizationDataset Image

Dataset Split

Train Set 87%
5254Images
Valid Set 9%
546Images
Test Set 4%
252Images

Preprocessing

Auto-Orient: Applied
Tile: 2 rows x 2 columns
Modify Classes: 0 remapped, 1 dropped
Filter Null: Require all images to contain annotations.

Augmentations

Outputs per training example: 2
Brightness: Between -20% and +20%
Blur: Up to 3.5px
Cutout: 5 boxes with 15% size each
Bounding Box: Flip: Horizontal, Vertical
Bounding Box: Rotation: Between -15° and +15°
Bounding Box: Brightness: Between -15% and +15%
Bounding Box: Noise: Up to 0.1% of pixels