Search papers, labs, and topics across Lattice.
The paper introduces TEMPO, a unified model designed to tackle the critical challenge of timestamping in large audio-language models (LALMs) by employing a supervised fine-tuning (SFT) approach enhanced with atomic timestamp tokens and a time-aware projector. This method significantly improves the model's ability to assign timestamps to audio events, outperforming existing state-of-the-art models like Audio Flamingo Next and Qwen3-Omni on a comprehensive benchmark of 10K samples. Additionally, the integration of reinforcement learning through GRPO serves as a refinement stage, yielding modest performance improvements beyond the initial SFT gains.
TEMPO not only achieves superior timestamping accuracy in audio-language models but also integrates innovative techniques like atomic timestamp tokens and reinforcement learning for enhanced performance.
Large audio-language models (LALMs) describe audio at the clip level but cannot assign timestamps to the events, speakers, or sounds they identify. Despite being essential for downstream tasks like speech recognition and dense audio captioning, timestamping remains a key limitation of most LALMs. We present TEMPO (Temporally-grounded Multi-task Post-training), the first unified model to handle audio, speech, and music timestamping tasks. Our core contribution is a supervised fine-tuning (SFT) stage built on three innovations: atomic timestamp tokens, a time-aware projector that injects sinusoidal wall-clock encodings into audio frame embeddings, and a distance-aware Gaussian loss. Our training is based on a synthetic-to-real curriculum. We further introduce, to our knowledge, the first application of reinforcement learning to unified audio timestamping, using GRPO with verifiable temporal rewards that directly optimize the evaluation objectives. Rather than serving as the primary source of performance gains, GRPO acts as a refinement stage on top of the SFT checkpoint, providing modest additional improvements. To support this work, we build a training dataset containing 119K samples and an evaluation benchmark containing 10K samples, drawn from established corpora across five tasks. On this benchmark, TEMPO outperforms Audio Flamingo Next and Qwen3-Omni, two state-of-the-art LALMs explicitly trained on timestamped data. Experiments confirm that SFT delivers most of these gains, with GRPO providing consistent but moderate refinements.
