Finetuning Strategies for Querying Sounds by Vocal Imitation

Source

• Source name: arXiv
• Source kind: paper
• Authors / organization: Aditya Bhattacharjee, Christos Plachouras, Sungkyun Chang, Emmanouil Benetos
• Venue: arXiv: cs.SD
• Published: 2026-08-19
• Collected: 2026-08-20
• URL: http://arxiv.org/abs/2608.19174v1
• PDF: https://arxiv.org/pdf/2608.19174v1
• DOI: UNKNOWN
• Registry group: academic_apis_and_preprints
• Evidence role: canonical_academic_index
• Language: en
• Country / region: UNKNOWN

Plain-Language Brief

The authors describe their winning strategy for a sound effect querying competition in 2025, which involved two fine-tuning strategies: contrastive learning with a frozen pretrained CED encoder and joint contrastive-triplet learning with semi-hard negatives using a MobileNetV3 encoder.

What the Document Says

• The authors submitted their strategy to the AES AIMLA 2025 Challenge on querying sound effects by vocal imitation. They used two fine-tuning strategies: contrastive learning with a frozen, pretrained CED encoder and joint contrastive-triplet learning with semi-hard negatives using a MobileNetV3 encoder.
• The report has been updated to include details released after the challenge.

Key Concepts Mentioned

• AES AIMLA 2025 Challenge
• querying sound effects
• vocal imitation
• fine-tuning strategies
• contrastive learning
• triplet learning
• CED encoder
• MobileNetV3 encoder

Claims Made by the Document

• The authors' strategy was successful in the AES AIMLA 2025 Challenge.

Evidence Used

• The authors describe their winning submission to the AES AIMLA 2025 Challenge.

Methods / Approach

UNKNOWN

Limitations Stated by the Authors

UNKNOWN

Limitations Noted by the Agent

Agent interpretation — not from the source.

UNKNOWN

Open Questions

UNKNOWN

Notes for Future Q&A

May include interpretation. Labeled where it does.

UNKNOWN

Raw Abstract / Excerpt

This technical report describes our winning submission to the AES AIMLA 2025 Challenge on querying sound effects by vocal imitation. We investigate two complementary fine-tuning strategies: contrastive learning with a frozen, pretrained CED encoder, and joint contrastive-triplet learning with semi-hard negatives using a MobileNetV3 encoder. This report has been updated for posterity to include details released after the challenge.