# Rank-based loss for learning hierarchical representations

arXiv: https://arxiv.org/abs/2110.05941 · HTML: https://mlsanity.com/paper/2110.05941/

## Key facts

- Cited by 4 papers in the map (data through 14 Oct 2026).
- Builds on 5 of its 14 references that are in the map.
- First posted on arXiv 11 Oct 2021.

## Builds on (most cited first, 5 of 5)

- [CNN Architectures for Large-Scale Audio Classification](https://mlsanity.com/paper/1609.09430/) (2016)
- [Neural Audio Synthesis of Musical Notes with WaveNet Autoencoders](https://mlsanity.com/paper/1704.01279/) (2017)
- [Leveraging Hierarchical Structures for Few-Shot Musical Instrument Recognition](https://mlsanity.com/paper/2107.07029/) (2021)
- [Deep Feature Embedding and Hierarchical Classification for Audio Scene Classification](https://mlsanity.com/paper/2002.04857/) (2020)
- [Hierarchical learning for DNN-based acoustic scene classification](https://mlsanity.com/paper/1607.03682/) (2016)

## Built on by (most cited first, 4 of 4)

- [Learning Structured Representations by Embedding Class Hierarchy with Fast Optimal Transport](https://mlsanity.com/paper/2410.03052/) (2024)
- [Acoustic identification of individual animals with hierarchical contrastive learning](https://mlsanity.com/paper/2409.08673/) (2024)
- [Hybrid Losses for Hierarchical Embedding Learning](https://mlsanity.com/paper/2501.12796/) (2025)
- [No Free Lunch from Audio Pretraining in Bioacoustics: A Benchmark Study of Embeddings](https://mlsanity.com/paper/2508.10230/) (2025)

## Abstract (arXiv)

Hierarchical taxonomies are common in many contexts, and they are a very natural structure humans use to organise information. In machine learning, the family of methods that use the 'extra' information is called hierarchical classification. However, applied to audio classification, this remains relatively unexplored. Here we focus on how to integrate the hierarchical information of a problem to learn embeddings representative of the hierarchical relationships. Previously, triplet loss has been proposed to address this problem, however it presents some issues like requiring the careful construction of the triplets, and being limited in the extent of hierarchical information it uses at each iteration. In this work we propose a rank based loss function that uses hierarchical information and translates this into a rank ordering of target distances between the examples. We show that rank based loss is suitable to learn hierarchical representations of the data. By testing on unseen fine level classes we show that this method is also capable of learning hierarchically correct representations of the new classes. Rank based loss has two promising aspects, it is generalisable to hierarchies with any number of levels, and is capable of dealing with data with incomplete hierarchical labels.