Metric Learning with Adaptive Density Discrimination

11/18/2015
by   Oren Rippel, et al.
0

Distance metric learning (DML) approaches learn a transformation to a representation space where distance is in correspondence with a predefined notion of similarity. While such models offer a number of compelling benefits, it has been difficult for these to compete with modern classification algorithms in performance and even in feature extraction. In this work, we propose a novel approach explicitly designed to address a number of subtle yet important issues which have stymied earlier DML algorithms. It maintains an explicit model of the distributions of the different classes in representation space. It then employs this knowledge to adaptively assess similarity, and achieve local discrimination by penalizing class distribution overlap. We demonstrate the effectiveness of this idea on several tasks. Our approach achieves state-of-the-art classification results on a number of fine-grained visual recognition datasets, surpassing the standard softmax classifier and outperforming triplet loss by a relative margin of 30-40 computational performance, it alleviates training inefficiencies in the traditional triplet loss, reaching the same error in 5-30 times fewer iterations. Beyond classification, we further validate the saliency of the learnt representations via their attribute concentration and hierarchy recovery properties, achieving 10-25 25-50

READ FULL TEXT

page 2

page 3

page 11

page 12

research
03/01/2017

Incorporating Intra-Class Variance to Fine-Grained Visual Recognition

Fine-grained visual recognition aims to capture discriminative character...
research
10/26/2022

Deep Metric Learning with Adaptive Margin and Adaptive Scale for Acoustic Word Discrimination

Many recent loss functions in deep metric learning are expressed with lo...
research
09/27/2019

A weakly supervised adaptive triplet loss for deep metric learning

We address the problem of distance metric learning in visual similarity ...
research
09/09/2019

Deep Metric Learning with Density Adaptivity

The problem of distance metric learning is mostly considered from the pe...
research
03/19/2021

Improving Image co-segmentation via Deep Metric Learning

Deep Metric Learning (DML) is helpful in computer vision tasks. In this ...
research
01/26/2023

Inspecting class hierarchies in classification-based metric learning models

Most classification models treat all misclassifications equally. However...
research
12/29/2017

Significance of Softmax-based Features in Comparison to Distance Metric Learning-based Features

The extraction of useful deep features is important for many computer vi...

Please sign up or login with your details

Forgot password? Click here to reset