ACVUBench: Audio-Centric Video Understanding Benchmark

Audio often serves as an auxiliary modality in video understanding tasks of audio-visual large language models (LLMs), merely assisting in the comprehension of visual information. However, a thorough understanding of videos significantly depends on auditory information, as audio offers critical context, emotional cues, and semantic meaning that visual data alone often lacks. This paper proposes an audio-centric video understanding benchmark (ACVUBench) to evaluate the video comprehension capabilities of multimodal LLMs with a particular focus on auditory information. Specifically, ACVUBench incorporates 2,662 videos spanning 18 different domains with rich auditory information, together with over 13k high-quality human annotated or validated question-answer pairs. Moreover, ACVUBench introduces a suite of carefully designed audio-centric tasks, holistically testing the understanding of both audio content and audio-visual interactions in videos. A thorough evaluation across a diverse range of open-source and proprietary multimodal LLMs is performed, followed by the analyses of deficiencies in audio-visual LLMs. Demos are available at https://github.com/lark-png/ACVUBench.

SimpleRL-Zoo: Investigating and Taming Zero Reinforcement Learning for Open Base Models in the Wild

DeepSeek-R1 has shown that long chain-of-thought (CoT) reasoning can naturally emerge through a simple reinforcement learning (RL) framework with rule-based rewards, where the training may directly start from the base models-a paradigm referred to as zero RL training. Most recent efforts to reproduce zero RL training have primarily focused on the Qwen2.5 model series, which may not be representative as we find the base models already exhibit strong instruction-following and self-reflection abilities. In this work, we investigate zero RL training across 10 diverse base models, spanning different families and sizes including LLama3-8B, Mistral-7B/24B, DeepSeek-Math-7B, Qwen2.5-math-7B, and all Qwen2.5 models from 0.5B to 32B. Leveraging several key design strategies-such as adjusting format reward and controlling query difficulty-we achieve substantial improvements in both reasoning accuracy and response length across most settings. However, by carefully monitoring the training dynamics, we observe that different base models exhibit distinct patterns during training. For instance, the increased response length does not always correlate with the emergence of certain cognitive behaviors such as verification (i.e., the "aha moment"). Notably, we observe the "aha moment" for the first time in small models not from the Qwen family. We share the key designs that enable successful zero RL training, along with our findings and practices. To facilitate further research, we open-source the code, models, and analysis tools.

Improving LLM Video Understanding with 16 Frames Per Second

Human vision is dynamic and continuous. However, in video understanding with multimodal large language models (LLMs), existing methods primarily rely on static features extracted from images sampled at a fixed low frame rate of frame-per-second (FPS) $\leqslant$2, leading to critical visual information loss. In this paper, we introduce F-16, the first multimodal LLM designed for high-frame-rate video understanding. By increasing the frame rate to 16 FPS and compressing visual tokens within each 1-second clip, F-16 efficiently captures dynamic visual features while preserving key semantic information. Experimental results demonstrate that higher frame rates considerably enhance video understanding across multiple benchmarks, providing a new approach to improving video LLMs beyond scaling model size or training data. F-16 achieves state-of-the-art performance among 7-billion-parameter video LLMs on both general and fine-grained video understanding benchmarks, such as Video-MME and TemporalBench. Furthermore, F-16 excels in complex spatiotemporal tasks, including high-speed sports analysis (\textit{e.g.}, basketball, football, gymnastics, and diving), outperforming SOTA proprietary visual models like GPT-4o and Gemini-1.5-pro. Additionally, we introduce a novel decoding method for F-16 that enables highly efficient low-frame-rate inference without requiring model retraining. Upon acceptance, we will release the source code, model checkpoints, and data.

Video Instruction Tuning With Synthetic Data

The development of video large multimodal models (LMMs) has been hindered by the difficulty of curating large amounts of high-quality raw data from the web. To address this, we propose an alternative approach by creating a high-quality synthetic dataset specifically for video instruction-following, namely LLaVA-Video-178K. This dataset includes key tasks such as detailed captioning, open-ended question-answering (QA), and multiple-choice QA. By training on this dataset, in combination with existing visual instruction tuning data, we introduce LLaVA-Video, a new video LMM. Our experiments demonstrate that LLaVA-Video achieves strong performance across various video benchmarks, highlighting the effectiveness of our dataset. We plan to release the dataset, its generation pipeline, and the model checkpoints.

LLaVA-NeXT-Interleave: Tackling Multi-image, Video, and 3D in Large Multimodal Models

Visual instruction tuning has made considerable strides in enhancing the capabilities of Large Multimodal Models (LMMs). However, existing open LMMs largely focus on single-image tasks, their applications to multi-image scenarios remains less explored. Additionally, prior LMM research separately tackles different scenarios, leaving it impossible to generalize cross scenarios with new emerging capabilities. To this end, we introduce LLaVA-NeXT-Interleave, which simultaneously tackles Multi-image, Multi-frame (video), Multi-view (3D), and Multi-patch (single-image) scenarios in LMMs. To enable these capabilities, we regard the interleaved data format as a general template and compile the M4-Instruct dataset with 1,177.6k samples, spanning 4 primary domains with 14 tasks and 41 datasets. We also curate the LLaVA-Interleave Bench to comprehensively evaluate the multi-image performance of LMMs. Through extensive experiments, LLaVA-NeXT-Interleave achieves leading results in multi-image, video, and 3D benchmarks, while maintaining the performance of single-image tasks. Besides, our model also exhibits several emerging capabilities, e.g., transferring tasks across different settings and modalities. Code is available at https://github.com/LLaVA-VL/LLaVA-NeXT

video-SALMONN: Speech-Enhanced Audio-Visual Large Language Models

Speech understanding as an element of the more generic video understanding using audio-visual large language models (av-LLMs) is a crucial yet understudied aspect. This paper proposes video-SALMONN, a single end-to-end av-LLM for video processing, which can understand not only visual frame sequences, audio events and music, but speech as well. To obtain fine-grained temporal information required by speech understanding, while keeping efficient for other video elements, this paper proposes a novel multi-resolution causal Q-Former (MRC Q-Former) structure to connect pre-trained audio-visual encoders and the backbone large language model. Moreover, dedicated training approaches including the diversity loss and the unpaired audio-visual mixed training scheme are proposed to avoid frames or modality dominance. On the introduced speech-audio-visual evaluation benchmark, video-SALMONN achieves more than 25\% absolute accuracy improvements on the video-QA task and over 30\% absolute accuracy improvements on audio-visual QA tasks with human speech. In addition, video-SALMONN demonstrates remarkable video comprehension and reasoning abilities on tasks that are unprecedented by other av-LLMs. Our training code and model checkpoints are available at \texttt{\url{https://github.com/bytedance/SALMONN/}}.

Can Large Language Models Understand Spatial Audio?

This paper explores enabling large language models (LLMs) to understand spatial information from multichannel audio, a skill currently lacking in auditory LLMs. By leveraging LLMs' advanced cognitive and inferential abilities, the aim is to enhance understanding of 3D environments via audio. We study 3 spatial audio tasks: sound source localization (SSL), far-field speech recognition (FSR), and localisation-informed speech extraction (LSE), achieving notable progress in each task. For SSL, our approach achieves an MAE of $2.70^{\circ}$ on the Spatial LibriSpeech dataset, substantially surpassing the prior benchmark of about $6.60^{\circ}$. Moreover, our model can employ spatial cues to improve FSR accuracy and execute LSE by selectively attending to sounds originating from a specified direction via text prompts, even amidst overlapping speech. These findings highlight the potential of adapting LLMs to grasp physical audio concepts, paving the way for LLM-based agents in 3D environments.

SA-SOT: Speaker-Aware Serialized Output Training for Multi-Talker ASR

Multi-talker automatic speech recognition plays a crucial role in scenarios involving multi-party interactions, such as meetings and conversations. Due to its inherent complexity, this task has been receiving increasing attention. Notably, the serialized output training (SOT) stands out among various approaches because of its simplistic architecture and exceptional performance. However, the frequent speaker changes in token-level SOT (t-SOT) present challenges for the autoregressive decoder in effectively utilizing context to predict output sequences. To address this issue, we introduce a masked t-SOT label, which serves as the cornerstone of an auxiliary training loss. Additionally, we utilize a speaker similarity matrix to refine the self-attention mechanism of the decoder. This strategic adjustment enhances contextual relationships within the same speaker's tokens while minimizing interactions between different speakers' tokens. We denote our method as speaker-aware SOT (SA-SOT). Experiments on the Librispeech datasets demonstrate that our SA-SOT obtains a relative cpWER reduction ranging from 12.75% to 22.03% on the multi-talker test sets. Furthermore, with more extensive training, our method achieves an impressive cpWER of 3.41%, establishing a new state-of-the-art result on the LibrispeechMix dataset.

SLIT: Boosting Audio-Text Pre-Training via Multi-Stage Learning and Instruction Tuning

Audio-text pre-training (ATP) has witnessed remarkable strides across a variety of downstream tasks. Yet, most existing pretrained audio models only specialize in either discriminative tasks or generative tasks. In this study, we develop SLIT, a novel ATP framework which transfers flexibly to both audio-text understanding and generation tasks, bootstrapping audio-text pre-training from frozen pretrained audio encoders and large language models. To bridge the modality gap during pre-training, we leverage Q-Former, which undergoes a multi-stage pre-training process. The first stage enhances audio-text representation learning from a frozen audio encoder, while the second stage boosts audio-to-text generative learning with a frozen language model. Furthermore, we introduce an ATP instruction tuning strategy, which enables flexible and informative feature extraction tailered to the given instructions for different tasks. Experiments show that SLIT achieves superior performances on a variety of audio-text understanding and generation tasks, and even demonstrates strong generalization capabilities when directly applied to zero-shot scenarios.

Real3D-Portrait: One-shot Realistic 3D Talking Portrait Synthesis

One-shot 3D talking portrait generation aims to reconstruct a 3D avatar from an unseen image, and then animate it with a reference video or audio to generate a talking portrait video. The existing methods fail to simultaneously achieve the goals of accurate 3D avatar reconstruction and stable talking face animation. Besides, while the existing works mainly focus on synthesizing the head part, it is also vital to generate natural torso and background segments to obtain a realistic talking portrait video. To address these limitations, we present Real3D-Potrait, a framework that (1) improves the one-shot 3D reconstruction power with a large image-to-plane model that distills 3D prior knowledge from a 3D face generative model; (2) facilitates accurate motion-conditioned animation with an efficient motion adapter; (3) synthesizes realistic video with natural torso movement and switchable background using a head-torso-background super-resolution model; and (4) supports one-shot audio-driven talking face generation with a generalizable audio-to-motion model. Extensive experiments show that Real3D-Portrait generalizes well to unseen identities and generates more realistic talking portrait videos compared to previous methods. Video samples and source code are available at https://real3dportrait.github.io .