Searcharxiv⌕ Search
Home/Archive/arXiv

arXiv · 1804.07297

Deep Layered Learning in MIR

Anders Elowsson
View original source ↗Read source document ↗Export metadata ↓

Abstract

Deep learning has boosted the performance of many music information retrieval (MIR) systems in recent years. Yet, the complex hierarchical arrangement of music makes end-to-end learning hard for some MIR tasks - a very deep and flexible processing chain is necessary to model some aspect of music audio. Representations involving tones, chords, and rhythm are fundamental building blocks of music. This paper discusses how these can be used as intermediate targets and priors in MIR to deal with structurally complex learning problems, with learning modules connected in a directed acyclic graph. It is suggested that this strategy for inference, referred to as deep layered learning (DLL), can help generalization by (1) - enforcing the validity and invariance of intermediate representations during processing, and by (2) - letting the inferred representations establish the musical organization to support higher-level invariant processing. A background to modular music processing is provided together with an overview of previous publications. Relevant concepts from information processing, such as pruning, skip connections, and performance supervision are reviewed within the context of DLL. A test is finally performed, showing how layered learning affects pitch tracking. It is indicated that especially offsets are easier to detect if guided by extracted framewise fundamental frequencies.

Explore related subjects

cs.SDeess.AS

Keep this discovery

BibTeXRIS

Anders Elowsson. 2018-04-18. Deep Layered Learning in MIR. https://arxiv.org/abs/1804.07297

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related papers

arXiv2026

Xiaomi-CocktailASR-1 Technical Report

Recently, large language model (LLM) based ASR models have achieved significant progress, yet they generally lack support for multi-speaker scenarios, where the cocktail party problem remains a critical bottleneck for further advancing ASR. Existing TS-ASR methods, including end-to-end architectures with speaker embeddings and latest LLM-based explorations suffer from degraded single-speaker performance and the inability to reject when the target speaker is absent. In this paper, we propose Xiaomi-CocktailASR-1, an LLM-based end-to-end TS-ASR architecture. By utilizing reference speech as voiceprint prompts, it directly transcribes the target speaker's speech without requiring speech separation. Xiaomi-CocktailASR-1 maintains competitive performance in single-speaker scenarios, comparable to mainstream ASR models. It also features a negative sample rejection capability, outputting empty text when the target speaker is absent from the mixed speech. Additionally, Xiaomi-CocktailASR-1 supports a Chain-of-Thought (CoT) reasoning mode to provide explicit reasoning steps. Extensive experiments on various synthetic and real-world multispeaker benchmarks demonstrate that Xiaomi-CocktailASR-1 achieves state-of-the-art performance, effectively addressing the cocktail party problem through a unified architecture that balances multispeaker and single-speaker recognition accuracy, along with rejection capability.

cs.SD↗
arXiv2026

EConv-TasNet: Efficient Conv-TasNet for Effective Speech Separation

Conv-TasNet has served as a strong baseline for time-domain speech separation, and many studies have extended it with advanced architectures such as dual-path networks, U-Nets, and attention mechanisms. However, these methods often introduce high computational cost and complexity, limiting their deployment in resource-constrained scenarios. To address this issue, we propose eConv-TasNet, an efficient variant of Conv-TasNet that improves both effectiveness and efficiency without relying on resource-intensive modules. The proposed model consists of a group-wise early-splitting (GES) module and a multi-group feature aggregation (MGFA) module. GES generates discriminative speaker embeddings at intermediate stages, while MGFA progressively aggregates these group-level representations for refined mask estimation. Experimental results show that eConv-TasNet reduces model size by 22.4%, accelerates inference by 18.9%, and improves SI-SNRi by 14.0%-28.0% across three public benchmarks. Moreover, it achieves competitive performance compared with state-of-the-art methods while requiring significantly fewer parameters and lower inference cost. These results demonstrate a favorable efficiency-effectiveness trade-off for edge deployment.

cs.SD↗