DeepAI AI Chat
Log In Sign Up

Robustness and Overfitting Behavior of Implicit Background Models

08/21/2020
by   Shirley Liu, et al.
7

In this paper, we examine the overfitting behavior of image classification models modified with Implicit Background Estimation (SCrIBE), which transforms them into weakly supervised segmentation models that provide spatial domain visualizations without affecting performance. Using the segmentation masks, we derive an overfit detection criterion that does not require testing labels. In addition, we assess the change in model performance, calibration, and segmentation masks after applying data augmentations as overfitting reduction measures and testing on various types of distorted images.

READ FULL TEXT

page 2

page 5

11/27/2018

CIAN: Cross-Image Affinity Net for Weakly Supervised Semantic Segmentation

Weakly supervised semantic segmentation based on image-level labels aims...
06/06/2017

Incorporating Network Built-in Priors in Weakly-supervised Semantic Segmentation

Pixel-level annotations are expensive and time consuming to obtain. Henc...
05/08/2019

Weakly Labeling the Antarctic: The Penguin Colony Case

Antarctic penguins are important ecological indicators -- especially in ...
12/12/2018

Weakly Supervised Instance Segmentation Using Hybrid Network

Weakly-supervised instance segmentation, which could greatly save labor ...
11/04/2019

Superpixel-Based Background Recovery from Multiple Images

In this paper, we propose an intuitive method to recover background from...