SV3.3B: A Sports Video Understanding Model for Action Recognition

• URL: http://arxiv.org/abs/2507.17844v1
• Date: Wed, 23 Jul 2025 18:11:39 GMT
• Title: SV3.3B: A Sports Video Understanding Model for Action Recognition
• Abstract summary: This paper addresses the challenge of automated sports video analysis.<n>Current approaches struggle to capture the nuanced biomechanical transitions essential for meaningful sports analysis.<n>We introduce SV3.3B, a lightweight 3.3B parameter video understanding model.
• Score: 0.0
• License: http://creativecommons.org/licenses/by-nc-nd/4.0/
• Abstract: This paper addresses the challenge of automated sports video analysis, which has traditionally been limited by computationally intensive models requiring server-side processing and lacking fine-grained understanding of athletic movements. Current approaches struggle to capture the nuanced biomechanical transitions essential for meaningful sports analysis, often missing critical phases like preparation, execution, and follow-through that occur within seconds. To address these limitations, we introduce SV3.3B, a lightweight 3.3B parameter video understanding model that combines novel temporal motion difference sampling with self-supervised learning for efficient on-device deployment. Our approach employs a DWT-VGG16-LDA based keyframe extraction mechanism that intelligently identifies the 16 most representative frames from sports sequences, followed by a V-DWT-JEPA2 encoder pretrained through mask-denoising objectives and an LLM decoder fine-tuned for sports action description generation. Evaluated on a subset of the NSVA basketball dataset, SV3.3B achieves superior performance across both traditional text generation metrics and sports-specific evaluation criteria, outperforming larger closed-source models including GPT-4o variants while maintaining significantly lower computational requirements. Our model demonstrates exceptional capability in generating technically detailed and analytically rich sports descriptions, achieving 29.2% improvement over GPT-4o in ground truth validation metrics, with substantial improvements in information density, action complexity, and measurement precision metrics essential for comprehensive athletic analysis. Model Available at https://huggingface.co/sportsvision/SV3.3B.

Related papers

• Digitizing Coaching Intelligence: An Agentic Framework for Holistic Athlete Profiling using VLM and RAG [2.059090414809123]
  This paper presents a novel, LLM-based hybrid agentic framework for automated, holistic athlete profiling.<n>To overcome the latency and token constraints associated with multimodal video processing, we introduce a 3 X 3 "Smart Grid" temporal chunking strategy.<n>We also implement a dual-persistence Retrieval-Augmented Generation pipeline utilizing a vector search engine (ChromaDB)
  arXiv  Detail & Related papers  (2026-06-26T19:53:26Z)
• SoccerLens: Grounded Soccer Video Understanding Beyond Accuracy [52.09081885158698]
  Vision-language models (VLMs) have recently shown strong potential in soccer video understanding.<n>Existing evaluation protocols focus on classification accuracy and do not assess visual grounding.<n>We introduce SoccerLens, a benchmark for grounded soccer video understanding.
  arXiv  Detail & Related papers  (2026-05-10T15:18:08Z)
• COVTrack++: Learning Open-Vocabulary Multi-Object Tracking from Continuous Videos via a Synergistic Paradigm [59.26203051651017]
  C-TAO is the first continuously annotated training set for Open-Vocabulary Multi-Object Tracking (OVMOT)<n>For the framework bottleneck, we propose COVTrack++, a synergistic framework that achieves a bidirectional reciprocal mechanism between detection and association through three modules.<n>Experiments on TAO demonstrate state-of-the-art performance, with novel TETA reaching 35.4% and 30.5% on validation and test sets, improving novel AssocA by 4.8% and novel LocA by 5.8% over previous methods.
  arXiv  Detail & Related papers  (2026-03-25T07:20:27Z)
• What Makes VLMs Robust? Towards Reconciling Robustness and Accuracy in Vision-Language Models [64.99746027349767]
  Adversarial Robustness Adaptation (R-Adapt) is a framework that freezes all pre-trained weights and introduces minimal, insight-driven adaptations only in the initial layers.<n>R-Adapt generalizes efficiently to large vision-language models (e.g., LLaVA and Qwen-VL) to enhance their robustness.
  arXiv  Detail & Related papers  (2026-03-13T09:02:11Z)
• Understanding vision transformer robustness through the lens of out-of-distribution detection [59.72757235382676]
  Quantization reduces memory and inference costs at the risk of performance loss.<n>We investigate the behaviour of quantized small-variant popular vision transformers (DeiT, DeiT3, and ViT) on common out-of-distribution (OOD) datasets.
  arXiv  Detail & Related papers  (2026-02-01T22:00:59Z)
• TimeLens: Rethinking Video Temporal Grounding with Multimodal LLMs [81.78017865436816]
  We present TimeLens, a systematic investigation into building MLLMs with strong video temporal grounding ability.<n>We first expose critical quality issues in existing VTG benchmarks and introduce TimeLens-Bench, comprising meticulously re-annotated versions of three popular benchmarks.<n>We also address noisy training data through an automated re-annotation pipeline, yielding TimeLens-100K, a large-scale, high-quality training dataset.
  arXiv  Detail & Related papers  (2025-12-16T18:59:58Z)
• SportsGPT: An LLM-driven Framework for Interpretable Sports Motion Assessment and Training Guidance [22.480385290370663]
  We propose SportsGPT, an LLM-driven framework for interpretable sports motion assessment and training guidance.<n> MotionDTW is a two-stage time series alignment algorithm designed for accurate extraction from skeleton-based motion sequences.<n>SportsRAG is a RAG-based training guidance model built upon Qwen3.
  arXiv  Detail & Related papers  (2025-12-16T06:05:55Z)
• Open-o3 Video: Grounded Video Reasoning with Explicit Spatio-Temporal Evidence [70.2803680525165]
  We introduce Open-o3 Video, a non-agent framework that integrates explicit evidence into video reasoning.<n>The model highlights key objects and bounding boxes alongside its answers, allowing reasoning to be grounded in concrete visual observations.<n>On V-STAR benchmark, Open-o3 Video achieves state-of-the-art performance, raising mAM by 14.4% and mL timestamp by 24.2%.
  arXiv  Detail & Related papers  (2025-10-23T14:05:56Z)
• Biomechanical-phase based Temporal Segmentation in Sports Videos: a Demonstration on Javelin-Throw [0.0]
  Traditional sports analytics techniques rely on manual annotation or laboratory-based instrumentation.<n>We present a novel unsupervised framework for such a contextually aware temporal segmentation.<n>Our approach achieves 71.02% mean average precision (mAP) and 74.61% F1-score on test data, substantially higher than competing baselines.
  arXiv  Detail & Related papers  (2025-09-29T11:11:46Z)
• CronusVLA: Transferring Latent Motion Across Time for Multi-Frame Prediction in Manipulation [67.1520483301709]
  CronusVLA is a unified framework that extends single-frame VLA models to the multi-frame paradigm through an efficient post-training stage.<n>CronusVLA achieves state-of-the-art performance on SimplerEnv with 70.9% success rate, and 12.7% improvement over OpenVLA on LIBERO.
  arXiv  Detail & Related papers  (2025-06-24T17:30:27Z)
• AthletePose3D: A Benchmark Dataset for 3D Human Pose Estimation and Kinematic Validation in Athletic Movements [4.653030985708889]
  AthletePose3D is a novel dataset designed to capture high-speed, high-acceleration athletic movements.<n>We evaluate state-of-the-art (SOTA) monocular 2D and 3D pose estimation models on the dataset.
  arXiv  Detail & Related papers  (2025-03-10T16:16:02Z)
• DSV: Exploiting Dynamic Sparsity to Accelerate Large-Scale Video DiT Training [85.04885553561164]
  Diffusion Transformers (DiTs) have shown remarkable performance in generating high-quality videos.<n>DiTs can consume up to 95% of processing time and demand specialized context parallelism.<n>This paper introduces DSV to accelerate video DiT training by leveraging the dynamic attention sparsity we empirically observe.
  arXiv  Detail & Related papers  (2025-02-11T14:39:59Z)
• SMI: An Information-Theoretic Metric for Predicting Model Knowledge Solely from Pre-Training Signals [51.60874286674908]
  We aim to predict performance in closed-book question answering (QA), a vital downstream task indicative of a model's internal knowledge.<n>We conduct large-scale retrieval and semantic analysis across the pre-training corpora of 21 publicly available and 3 custom-trained large language models.<n>Building on these foundations, we propose Size-dependent Mutual Information (SMI), an information-theoretic metric that linearly correlates pre-training data characteristics, model size, and QA accuracy.
  arXiv  Detail & Related papers  (2025-02-06T13:23:53Z)
• Enhanced Multi-Object Tracking Using Pose-based Virtual Markers in 3x3 Basketball [3.072051066949152]
  We propose a novel virtual marker (VM) MOT method for team sports, named Sports-vmTracking.<n>Our approach achieved an average HOTA score of 72.3%, over 10 points higher than other state-of-theart methods without VM, and resulted in 0 ID switches.
  arXiv  Detail & Related papers  (2024-12-09T07:16:50Z)
• AIGV-Assessor: Benchmarking and Evaluating the Perceptual Quality of Text-to-Video Generation with LMM [54.44479359918971]
  We first present AIGVQA-DB, a large-scale dataset comprising 36,576 AIGVs generated by 15 advanced text-to-video models using 1,048 prompts. We then introduce AIGV-Assessor, a novel VQA model that leverages intricate quality attributes to capture precise video quality scores and pair video preferences.
  arXiv  Detail & Related papers  (2024-11-26T08:43:15Z)
• ALOcc: Adaptive Lifting-based 3D Semantic Occupancy and Cost Volume-based Flow Prediction [89.89610257714006]
  Existing methods prioritize higher accuracy to cater to the demands of these tasks. We introduce a series of targeted improvements for 3D semantic occupancy prediction and flow estimation. Our purelytemporalal architecture framework, named ALOcc, achieves an optimal tradeoff between speed and accuracy.
  arXiv  Detail & Related papers  (2024-11-12T11:32:56Z)
• Unlabeled Action Quality Assessment Based on Multi-dimensional Adaptive Constrained Dynamic Time Warping [12.639728404278255]
  This paper presents an unlabeled Multi-Dimensional Exercise Distance Adaptive Constrained Dynamic Time Warping (MED-ACDTW) method for action quality assessment. Our approach uses both 2D and 3D spatial dimensions, along with multiple human body features, to compare features from template and test videos. The adaptive constraint scheme enhances the discriminability of action quality assessment by approximately 30%.
  arXiv  Detail & Related papers  (2024-10-18T04:00:26Z)
• EITNet: An IoT-Enhanced Framework for Real-Time Basketball Action Recognition [17.068932442773864]
  EITNet is a framework that combines deep learning, I3Dtemporal object extraction, and TimeSformer for temporal analysis. Our contributions include developing a robust architecture that improves recognition accuracy to 92%. The integration of IoT technology enhances real-time data processing, providing adaptive insights into player performance and strategy.
  arXiv  Detail & Related papers  (2024-10-13T18:21:15Z)
• Temporal Reasoning Transfer from Text to Video [51.68487044397409]
  Video Large Language Models (Video LLMs) struggle with tracking temporal changes and reasoning about temporal relationships. We introduce the Textual Temporal reasoning Transfer (T3) to transfer temporal reasoning abilities from text to video domains. LongVA-7B model achieves competitive performance on comprehensive video benchmarks.
  arXiv  Detail & Related papers  (2024-10-08T16:10:29Z)
• S^2Former-OR: Single-Stage Bi-Modal Transformer for Scene Graph Generation in OR [50.435592120607815]
  Scene graph generation (SGG) of surgical procedures is crucial in enhancing holistically cognitive intelligence in the operating room (OR) Previous works have primarily relied on multi-stage learning, where the generated semantic scene graphs depend on intermediate processes with pose estimation and object detection. In this study, we introduce a novel single-stage bi-modal transformer framework for SGG in the OR, termed S2Former-OR.
  arXiv  Detail & Related papers  (2024-02-22T11:40:49Z)
• Towards Active Learning for Action Spotting in Association Football Videos [59.84375958757395]
  Analyzing football videos is challenging and requires identifying subtle and diverse-temporal patterns. Current algorithms face significant challenges when learning from limited annotated data. We propose an active learning framework that selects the most informative video samples to be annotated next.
  arXiv  Detail & Related papers  (2023-04-09T11:50:41Z)
• Sports Video Analysis on Large-Scale Data [10.24207108909385]
  This paper investigates the modeling of automated machine description on sports video. We propose a novel large-scale NBA dataset for Sports Video Analysis (NSVA) with a focus on captioning.
  arXiv  Detail & Related papers  (2022-08-09T16:59:24Z)

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.