Audio Flamingo 2: An Audio-Language Model with Long-Audio Understanding and Expert Reasoning Abilities

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Ghosh, Sreyan, Kong, Zhifeng, Kumar, Sonal, Sakshi, S, Kim, Jaehyeon, Ping, Wei, Valle, Rafael, Manocha, Dinesh, Catanzaro, Bryan
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866909527074406400
author Ghosh, Sreyan
Kong, Zhifeng
Kumar, Sonal
Sakshi, S
Kim, Jaehyeon
Ping, Wei
Valle, Rafael
Manocha, Dinesh
Catanzaro, Bryan
author_facet Ghosh, Sreyan
Kong, Zhifeng
Kumar, Sonal
Sakshi, S
Kim, Jaehyeon
Ping, Wei
Valle, Rafael
Manocha, Dinesh
Catanzaro, Bryan
contents Understanding and reasoning over non-speech sounds and music are crucial for both humans and AI agents to interact effectively with their environments. In this paper, we introduce Audio Flamingo 2 (AF2), an Audio-Language Model (ALM) with advanced audio understanding and reasoning capabilities. AF2 leverages (i) a custom CLAP model, (ii) synthetic Audio QA data for fine-grained audio reasoning, and (iii) a multi-stage curriculum learning strategy. AF2 achieves state-of-the-art performance with only a 3B parameter small language model, surpassing large open-source and proprietary models across over 20 benchmarks. Next, for the first time, we extend audio understanding to long audio segments (30 secs to 5 mins) and propose LongAudio, a large and novel dataset for training ALMs on long audio captioning and question-answering tasks. Fine-tuning AF2 on LongAudio leads to exceptional performance on our proposed LongAudioBench, an expert annotated benchmark for evaluating ALMs on long audio understanding capabilities. We conduct extensive ablation studies to confirm the efficacy of our approach. Project Website: https://research.nvidia.com/labs/adlr/AF2/.
format Preprint
id arxiv_https___arxiv_org_abs_2503_03983
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Audio Flamingo 2: An Audio-Language Model with Long-Audio Understanding and Expert Reasoning Abilities
Ghosh, Sreyan
Kong, Zhifeng
Kumar, Sonal
Sakshi, S
Kim, Jaehyeon
Ping, Wei
Valle, Rafael
Manocha, Dinesh
Catanzaro, Bryan
Sound
Computation and Language
Machine Learning
Audio and Speech Processing
Understanding and reasoning over non-speech sounds and music are crucial for both humans and AI agents to interact effectively with their environments. In this paper, we introduce Audio Flamingo 2 (AF2), an Audio-Language Model (ALM) with advanced audio understanding and reasoning capabilities. AF2 leverages (i) a custom CLAP model, (ii) synthetic Audio QA data for fine-grained audio reasoning, and (iii) a multi-stage curriculum learning strategy. AF2 achieves state-of-the-art performance with only a 3B parameter small language model, surpassing large open-source and proprietary models across over 20 benchmarks. Next, for the first time, we extend audio understanding to long audio segments (30 secs to 5 mins) and propose LongAudio, a large and novel dataset for training ALMs on long audio captioning and question-answering tasks. Fine-tuning AF2 on LongAudio leads to exceptional performance on our proposed LongAudioBench, an expert annotated benchmark for evaluating ALMs on long audio understanding capabilities. We conduct extensive ablation studies to confirm the efficacy of our approach. Project Website: https://research.nvidia.com/labs/adlr/AF2/.
title Audio Flamingo 2: An Audio-Language Model with Long-Audio Understanding and Expert Reasoning Abilities
topic Sound
Computation and Language
Machine Learning
Audio and Speech Processing
url https://arxiv.org/abs/2503.03983