Sparrow: Data-Efficient Video-LLM with Text-to-Image Augmentation

• URL: http://arxiv.org/abs/2411.19951v5
• Date: Tue, 22 Jul 2025 12:09:51 GMT
• Title: Sparrow: Data-Efficient Video-LLM with Text-to-Image Augmentation
• Authors: Shukang Yin, Chaoyou Fu, Sirui Zhao, Chunjiang Ge, Yan Yang, Yuhan Dai, Yongdong Luo, Tong Xu, Caifeng Shan, Enhong Chen,
• Abstract summary: This work revisits scaling with synthetic data and focuses on developing video-LLMs from a data-centric perspective.<n>We propose a data augmentation method called Sparrow, which synthesizes video-like samples from pure text instruction data.<n>Our proposed method achieves performance comparable to or even superior to that of baselines trained with significantly more samples.
• Score: 57.34255010956452
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Recent years have seen the success of Multimodal Large Language Models (MLLMs) in the domain of vision understanding. The success of these models can largely be attributed to the dominant scaling law, which states that larger parameter sizes and data volumes contribute to better performance. Notably, data scaling has been primarily driven by automatic data pipelines, which focus on the self-instruction of LLMs. The paradigm has been taken for granted for quite some time, but the study of the effectiveness of scaling with these data has been neglected for a long time. In this context, this work revisits scaling with synthetic data and focuses on developing video-LLMs from a data-centric perspective. Our primary study approach involves fine-tuning pre-trained image-LLMs with video data and examining learning efficiency through data scaling. Results from our preliminary experiments reveal a low learning efficiency phenomenon when simply scaling up video data samples, which, through our probing, can be ascribed to a lack of instruction diversity. Aiming at this issue, we propose a data augmentation method called Sparrow, which synthesizes video-like samples from pure text instruction data. Mixing these synthetic samples with the video data enables a more efficient training scheme. Through comprehensive experiments, we demonstrate that our proposed method achieves performance comparable to or even superior to that of baselines trained with significantly more samples. Meanwhile, we find that incorporating these synthetic samples can enhance the performance of long video understanding without requiring training on long video data. The code and data examples are available at https://github.com/VITA-MLLM/Sparrow.

Related papers

• Trust the Model: Compact VLMs as In-Context Judges for Image-Text Data Quality [5.750869893508341]
  Vision-language models (VLMs) extend the conventional large language models by integrating visual data, enabling richer multimodal reasoning.<n>We introduce a streamlined data filtration framework that employs a compact VLM, fine-tuned on a high-quality image-caption annotated dataset.<n>This model effectively evaluates and filters potential training samples based on caption and image quality and alignment.
  arXiv  Detail & Related papers  (2025-07-27T07:20:25Z)
• Video Dataset Condensation with Diffusion Models [7.44997213284633]
  Video dataset distillation is a promising solution to generate a compact synthetic dataset that retains the essential information from a large real dataset.<n>In this paper, we focus on video dataset distillation by employing a video diffusion model to generate high-quality synthetic videos.<n>To enhance representativeness, we introduce Video Spatio-Temporal U-Net (VST-UNet), a model designed to select a diverse and informative subset of videos.<n>We validate the effectiveness of our approach through extensive experiments on four benchmark datasets, demonstrating performance improvements of up to (10.61%) over the state-of-the
  arXiv  Detail & Related papers  (2025-05-10T15:12:19Z)
• PerceptionLM: Open-Access Data and Models for Detailed Visual Understanding [126.15907330726067]
  We build a Perception Model Language (PLM) in a fully open and reproducible framework for transparent research in image and video understanding. We analyze standard training pipelines without distillation from models and explore large-scale synthetic data to identify critical data gaps.
  arXiv  Detail & Related papers  (2025-04-17T17:59:56Z)
• Transferable text data distillation by trajectory matching [27.826518926355295]
  The data distillation method aims to synthesize a small number of data samples to achieve the training effect of the full data set. In this work, we proposed a method that involves learning pseudo prompt data based on trajectory matching. Evaluations on two benchmarks, including ARC-Easy and MMLU instruction tuning datasets, established the superiority of our distillation approach over the SOTA data selection method LESS.
  arXiv  Detail & Related papers  (2025-04-14T02:39:26Z)
• AccVideo: Accelerating Video Diffusion Model with Synthetic Dataset [55.82208863521353]
  We propose AccVideo to reduce the inference steps for accelerating video diffusion models with synthetic dataset.<n>Our model achieves 8.5x improvements in generation speed compared to the teacher model.<n>Compared to previous accelerating methods, our approach is capable of generating videos with higher quality and resolution.
  arXiv  Detail & Related papers  (2025-03-25T08:52:07Z)
• Few-shot LLM Synthetic Data with Distribution Matching [37.55363714371521]
  Large language models (LLMs) produce high-quality synthetic data to enhance the performance of smaller models. LLMs-generated synthetic data often differs from the real data in key language attributes. We introduce SynAlign: a synthetic data generation and filtering framework based on key attribute distribution matching.
  arXiv  Detail & Related papers  (2025-02-09T16:43:32Z)
• Free Video-LLM: Prompt-guided Visual Perception for Efficient Training-free Video LLMs [56.040198387038025]
  We present a novel prompt-guided visual perception framework (abbreviated as Free Video-LLM) for efficient inference of training-free video LLMs. Our method effectively reduces the number of visual tokens while maintaining high performance across multiple video question-answering benchmarks.
  arXiv  Detail & Related papers  (2024-10-14T12:35:12Z)
• Video Instruction Tuning With Synthetic Data [84.64519990333406]
  We create 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.
  arXiv  Detail & Related papers  (2024-10-03T17:36:49Z)
• Entropy Law: The Story Behind Data Compression and LLM Performance [115.70395740286422]
  We find that model performance is negatively correlated to the compression ratio of training data, which usually yields a lower training loss. Based on the findings of the entropy law, we propose a quite efficient and universal data selection method. We also present an interesting application of entropy law that can detect potential performance risks at the beginning of model training.
  arXiv  Detail & Related papers  (2024-07-09T08:14:29Z)
• Auto Cherry-Picker: Learning from High-quality Generative Data Driven by Language [41.40908753726324]
  Diffusion models can generate realistic and diverse images, potentially facilitating data availability for data-intensive perception tasks.<n>We present textbfAuto textbfCherry-textbfPicker (ACP), a novel framework that generates high-quality cross-modality training samples at scale.
  arXiv  Detail & Related papers  (2024-06-28T17:53:18Z)
• Learning Semantic Proxies from Visual Prompts for Parameter-Efficient Fine-Tuning in Deep Metric Learning [13.964106147449051]
  Existing solutions concentrate on fine-tuning the pre-trained models on conventional image datasets. We propose a novel and effective framework based on learning Visual Prompts (VPT) in the pre-trained Vision Transformers (ViT) We demonstrate that our new approximations with semantic information are superior to representative capabilities.
  arXiv  Detail & Related papers  (2024-02-04T04:42:05Z)
• Harvest Video Foundation Models via Efficient Post-Pretraining [67.30842563833185]
  We propose an efficient framework to harvest video foundation models from image ones. Our method is intuitively simple by randomly dropping input video patches and masking out input text during the post-pretraining procedure. Our method achieves state-of-the-art performances, which are comparable to some heavily pretrained video foundation models.
  arXiv  Detail & Related papers  (2023-10-30T14:06:16Z)
• Weakly Supervised Two-Stage Training Scheme for Deep Video Fight Detection Model [0.0]
  Fight detection in videos is an emerging deep learning application with today's prevalence of surveillance systems and streaming media. Previous work has largely relied on action recognition techniques to tackle this problem. We design the fight detection model as a composition of an action-aware feature extractor and an anomaly score generator.
  arXiv  Detail & Related papers  (2022-09-23T08:29:16Z)
• Cross-modal Manifold Cutmix for Self-supervised Video Representation Learning [50.544635516455116]
  This paper focuses on designing video augmentation for self-supervised learning. We first analyze the best strategy to mix videos to create a new augmented video sample. We propose Cross-Modal Manifold Cutmix (CMMC) that inserts a video tesseract into another video tesseract in the feature space across two different modalities.
  arXiv  Detail & Related papers  (2021-12-07T18:58:33Z)
• VALUE: A Multi-Task Benchmark for Video-and-Language Understanding Evaluation [124.02278735049235]
  VALUE benchmark aims to cover a broad range of video genres, video lengths, data volumes, and task difficulty levels. We evaluate various baseline methods with and without large-scale VidL pre-training. The significant gap between our best model and human performance calls for future study for advanced VidL models.
  arXiv  Detail & Related papers  (2021-06-08T18:34:21Z)
• Automatic Curation of Large-Scale Datasets for Audio-Visual Representation Learning [62.47593143542552]
  We describe a subset optimization approach for automatic dataset curation. We demonstrate that our approach finds videos with high audio-visual correspondence and show that self-supervised models trained on our data, despite being automatically constructed, achieve similar downstream performances to existing video datasets with similar scales.
  arXiv  Detail & Related papers  (2021-01-26T14:27:47Z)
• Omni-supervised Facial Expression Recognition via Distilled Data [120.11782405714234]
  We propose omni-supervised learning to exploit reliable samples in a large amount of unlabeled data for network training. We experimentally verify that the new dataset can significantly improve the ability of the learned FER model. To tackle this, we propose to apply a dataset distillation strategy to compress the created dataset into several informative class-wise images.
  arXiv  Detail & Related papers  (2020-05-18T09:36:51Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.