# Merlin L48 Spectrogram Dataset

arXiv: https://arxiv.org/abs/2511.00252 · HTML: https://mlsanity.com/paper/2511.00252/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 22 of its 42 references that are in the map.
- First posted on arXiv 31 Oct 2025.

## Builds on (most cited first, 22 of 22)

- [Deep Residual Learning for Image Recognition](https://mlsanity.com/paper/1512.03385/) (2015)
- [Learning Transferable Visual Models From Natural Language Supervision](https://mlsanity.com/paper/2103.00020/) (2021)
- [Microsoft COCO: Common Objects in Context](https://mlsanity.com/paper/1405.0312/) (2014)
- [Rethinking the Inception Architecture for Computer Vision](https://mlsanity.com/paper/1512.00567/) (2015)
- [Visual Genome: Connecting Language and Vision Using Crowdsourced Dense Image Annotations](https://mlsanity.com/paper/1602.07332/) (2016)
- [General Multi-label Image Classification with Transformers](https://mlsanity.com/paper/2011.14027/) (2020)
- [Towards a Visual Privacy Advisor: Understanding and Predicting Privacy Risks in Images](https://mlsanity.com/paper/1703.10660/) (2017)
- [Multi-Label Learning from Single Positive Labels](https://mlsanity.com/paper/2106.09708/) (2021)
- [Spatial Implicit Neural Representations for Global-Scale Species Mapping](https://mlsanity.com/paper/2306.02564/) (2023)
- [Multi-Label Learning with Global and Local Label Correlation](https://mlsanity.com/paper/1704.01415/) (2017)
- [Large Loss Matters in Weakly Supervised Multi-Label Classification](https://mlsanity.com/paper/2206.03740/) (2022)
- [Simple and Robust Loss Design for Multi-Label Learning with Missing Labels](https://mlsanity.com/paper/2112.07368/) (2021)
- [Source-free Video Domain Adaptation by Learning Temporal Consistency for Action Recognition](https://mlsanity.com/paper/2203.04559/) (2022)
- [Acknowledging the Unknown for Multi-label Learning with Single Positive Labels](https://mlsanity.com/paper/2203.16219/) (2022)
- [One Positive Label is Sufficient: Single-Positive Multi-Label Learning with Label Enhancement](https://mlsanity.com/paper/2206.00517/) (2022)
- [Spatial Consistency Loss for Training Multi-Label Classifiers from Single-Label Annotations](https://mlsanity.com/paper/2203.06127/) (2022)
- [The iNaturalist Sounds Dataset](https://mlsanity.com/paper/2506.00343/) (2025)
- [NIPS4Bplus: a richly annotated birdsong audio dataset](https://mlsanity.com/paper/1811.02275/) (2018)
- [Exploring Fine-Grained Audiovisual Categorization with the SSW60 Dataset](https://mlsanity.com/paper/2207.10664/) (2022)
- [G2NetPL: Generic Game-Theoretic Network for Partial-Label Image Classification](https://mlsanity.com/paper/2210.11469/) (2022)
- [PLMCL: Partial-Label Momentum Curriculum Learning for Multi-Label Image Classification](https://mlsanity.com/paper/2208.09999/) (2022)
- [Understanding Label Bias in Single Positive Multi-Label Learning](https://mlsanity.com/paper/2305.15584/) (2023)

## Built on by (most cited first, 1 of 1)

- [Masked Autoencoders with Limited Data: Does It Work? A Fine-Grained Bioacoustics Case Study](https://mlsanity.com/paper/2605.14031/) (2026)

## Abstract (arXiv)

In the single-positive multi-label (SPML) setting, each image in a dataset is labeled with the presence of a single class, while the true presence of other classes remains unknown. The challenge is to narrow the performance gap between this partially-labeled setting and fully-supervised learning, which often requires a significant annotation budget. Prior SPML methods were developed and benchmarked on synthetic datasets created by randomly sampling single positive labels from fully-annotated datasets like Pascal VOC, COCO, NUS-WIDE, and CUB200. However, this synthetic approach does not reflect real-world scenarios and fails to capture the fine-grained complexities that can lead to difficult misclassifications. In this work, we introduce the L48 dataset, a fine-grained, real-world multi-label dataset derived from recordings of bird sounds. L48 provides a natural SPML setting with single-positive annotations on a challenging, fine-grained domain, as well as two extended settings in which domain priors give access to additional negative labels. We benchmark existing SPML methods on L48 and observe significant performance differences compared to synthetic datasets and analyze method weaknesses, underscoring the need for more realistic and difficult benchmarks.