# Boosting Standard Classification Architectures Through a Ranking Regularizer

## TL;DR

This paper introduces a triplet loss regularizer to enhance standard classification architectures like ResNet and Inception, improving performance on fine-grained and imbalanced datasets without extra inference cost.

## Contribution

It demonstrates how to effectively incorporate triplet loss into standard models, overcoming previous computational challenges and supporting both classification and embedding tasks.

## Key findings

- Steady improvement on five fine-grained datasets
- Significant gains on an imbalanced video dataset
- Supports both classification and embedding without extra inference cost

## Abstract

We employ triplet loss as a feature embedding regularizer to boost classification performance. Standard architectures, like ResNet and Inception, are extended to support both losses with minimal hyper-parameter tuning. This promotes generality while fine-tuning pretrained networks. Triplet loss is a powerful surrogate for recently proposed embedding regularizers. Yet, it is avoided due to large batch-size requirement and high computational cost. Through our experiments, we re-assess these assumptions. During inference, our network supports both classification and embedding tasks without any computational overhead. Quantitative evaluation highlights a steady improvement on five fine-grained recognition datasets. Further evaluation on an imbalanced video dataset achieves significant improvement. Triplet loss brings feature embedding characteristics like nearest neighbor to classification models. Code available at \url{http://bit.ly/2LNYEqL}.