# SPECTRA: Subspace-Preserving Embedding Calibration, Transport, and Replay for Fully Few-Shot Class-Incremental Audio Classification

arXiv: https://arxiv.org/abs/2608.25054 · HTML: https://mlsanity.com/paper/2608.25054/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 16 of its 25 references that are in the map.
- First posted on arXiv 25 Aug 2026.

## Builds on (most cited first, 16 of 16)

- [Sinkhorn Distances: Lightspeed Computation of Optimal Transportation Distances](https://mlsanity.com/paper/1306.0895/) (2013)
- [Learning to Prompt for Vision-Language Models](https://mlsanity.com/paper/2109.01134/) (2021)
- [Conditional Prompt Learning for Vision-Language Models](https://mlsanity.com/paper/2203.05557/) (2022)
- [Continual Learning with Deep Generative Replay](https://mlsanity.com/paper/1705.08690/) (2017)
- [Few-Shot Class-Incremental Learning](https://mlsanity.com/paper/2004.10956/) (2020)
- [Neural Audio Synthesis of Musical Notes with WaveNet Autoencoders](https://mlsanity.com/paper/1704.01279/) (2017)
- [Pengi: An Audio Language Model for Audio Tasks](https://mlsanity.com/paper/2305.11834/) (2023)
- [Few-Shot Incremental Learning with Continually Evolved Classifiers](https://mlsanity.com/paper/2104.03047/) (2021)
- [Forward Compatible Few-Shot Class-Incremental Learning](https://mlsanity.com/paper/2203.06953/) (2022)
- [Leveraging the Feature Distribution in Transfer-based Few-Shot Learning](https://mlsanity.com/paper/2006.03806/) (2020)
- [Who calls the shots? Rethinking Few-Shot Learning for Audio](https://mlsanity.com/paper/2110.09600/) (2021)
- [Few-shot Class-incremental Audio Classification Using Dynamically Expanded Classifier with Self-attention Modified Prototypes](https://mlsanity.com/paper/2305.19539/) (2023)
- [PALM: Few-Shot Prompt Learning for Audio Language Models](https://mlsanity.com/paper/2409.19806/) (2024)
- [Few-shot Class-incremental Audio Classification Using Adaptively-refined Prototypes](https://mlsanity.com/paper/2305.18045/) (2023)
- [Fully Few-shot Class-incremental Audio Classification Using Expandable Dual-embedding Extractor](https://mlsanity.com/paper/2406.08122/) (2024)
- [SpurAudio: A Benchmark for Studying Shortcut Learning in Few-Shot Audio Classification](https://mlsanity.com/paper/2605.13672/) (2026)

## Abstract (arXiv)

Fully few-shot class-incremental audio classification (FFCAC) requires recognizing new sound classes from only a handful of labeled examples per session, without forgetting previously learned classes and without any large base dataset. Existing methods typically freeze a pre-trained audio--language encoder and classify with point prototypes, but they suffer from significant performance degradation throughout the sessions due to generic feature representations. We propose SPECTRA, a framework built on a frozen encoder which adds three components. (i) a lightweight trainable adapter that calibrates the generic embeddings to the task; (ii) subspace feature replay, an exemplar-free anti-forgetting scheme that replays old classes by sampling from the low-rank subspace of their stored features; and (iii) a transductive optimal-transport refinement of prototypes at test time. Our central finding is that the subspace structure of the replay diminishes forgetting and outperforms naive Gaussian replay of equal variance. On three FFCAC benchmarks (NSynth-100, FSC-89, LS-100), SPECTRA improves average accuracy and reduces forgetting over current state-of-the-art methods, and our ablations statistically validate each component.