mil-benchmarks: Standardized Evaluation of Deep Multiple-Instance Learning Techniques

05/04/2021
by   Daniel Grahn, et al.
0

Multiple-instance learning is a subset of weakly supervised learning where labels are applied to sets of instances rather than the instances themselves. Under the standard assumption, a set is positive only there is if at least one instance in the set which is positive. This paper introduces a series of multiple-instance learning benchmarks generated from MNIST, Fashion-MNIST, and CIFAR10. These benchmarks test the standard, presence, absence, and complex assumptions and provide a framework for future benchmarks to be distributed. I implement and evaluate several multiple-instance learning techniques against the benchmarks. Further, I evaluate the Noisy-And method with label noise and find mixed results with different datasets. The models are implemented in TensorFlow 2.4.1 and are available on GitHub. The benchmarks are available from PyPi as mil-benchmarks and on GitHub.

READ FULL TEXT
research
05/29/2019

Probabilistic Decoupling of Labels in Classification

We investigate probabilistic decoupling of labels supplied for training,...
research
09/26/2013

Generative Multiple-Instance Learning Models For Quantitative Electromyography

We present a comprehensive study of the use of generative modeling appro...
research
06/18/2023

ProMIL: Probabilistic Multiple Instance Learning for Medical Imaging

Multiple Instance Learning (MIL) is a weakly-supervised problem in which...
research
11/01/2021

Nested Multiple Instance Learning with Attention Mechanisms

Multiple instance learning (MIL) is a type of weakly supervised learning...
research
06/23/2023

On Learning Latent Models with Multi-Instance Weak Supervision

We consider a weakly supervised learning scenario where the supervision ...
research
06/27/2012

A Convex Relaxation for Weakly Supervised Classifiers

This paper introduces a general multi-class approach to weakly supervise...
research
05/22/2022

Human Instance Matting via Mutual Guidance and Multi-Instance Refinement

This paper introduces a new matting task called human instance matting (...

Please sign up or login with your details

Forgot password? Click here to reset