Vipul Baghel

Maryam Viqar, Violeta Madjarova, Vipul Baghel et al.

Human veins are important for carrying the blood from the body-parts to the heart. The improper functioning of the human veins may arise from several venous diseases. Varicose vein is one such disease wherein back flow of blood can occur, often resulting in increased venous pressure or restricted blood flow due to changes in the structure of vein. To examine the functional characteristics of the varicose vein, it is crucial to study the physical and bio mechanical properties of the vein. This work proposes a segmentation model Opto-UNet, for segmenting the venous wall structure. Optical Coherence Tomography system is used to acquire images of varicose vein. As the extracted vein is not uniform in shape, hence adequate method of segmentation is required to segment the venous wall. Opto-UNet model is based on the U-Net architecture wherein a new block is integrated into the architecture, employing atrous and separable convolution to extract spatially wide-range and separable features maps for attaining advanced performance. Furthermore, the depth wise separable convolution significantly reduces the complexity of the network by optimizing the number of parameters. The model achieves accuracy of 0.9830, sensitivity of 0.8425 and specificity of 0.9980 using 8.54 million number of parameters. These results indicate that model is highly adequate in segmenting the varicose vein wall without deteriorating the segmentation quality along with reduced complexity

A. S. Karthikeyan, Vipul Baghel, Anish Monsley Kirupakaran et al.

Continuous tracking of boxers across multiple training sessions helps quantify traits required for the well-known ten-point-must system. However, continuous tracking of multiple athletes across multiple training sessions remains a challenge, because it is difficult to precisely segment bout boundaries in a recorded video stream. Furthermore, re-identification of the same athlete over different period or even within the same bout remains a challenge. Difficulties are further compounded when a single fixed view video is captured in top-view. This work summarizes our progress in creating a system in an economically single fixed top-view camera. Specifically, we describe improved algorithm for bout transition detection and in-bout continuous player identification without erroneous ID updation or ID switching. From our custom collected data of ~11 hours (athlete count: 45, bouts: 189), our transition detection algorithm achieves 90% accuracy and continuous ID tracking achieves IDU=0, IDS=0.

Rahul Kumar, Vipul Baghel, Sudhanshu Singh et al.

Accurate analysis of combat sports using computer vision has gained traction in recent years, yet the development of robust datasets remains a major bottleneck due to the dynamic, unstructured nature of actions and variations in recording environments. In this work, we present a comprehensive, well-annotated video dataset tailored for punch detection and classification in boxing. The dataset comprises 6,915 high-quality punch clips categorized into six distinct punch types, extracted from 20 publicly available YouTube sparring sessions and involving 18 different athletes. Each clip is manually segmented and labeled to ensure precise temporal boundaries and class consistency, capturing a wide range of motion styles, camera angles, and athlete physiques. This dataset is specifically curated to support research in real-time vision-based action recognition, especially in low-resource and unconstrained environments. By providing a rich benchmark with diverse punch examples, this contribution aims to accelerate progress in movement analysis, automated coaching, and performance assessment within boxing and related domains.

Bikash Kumar Badatya, Vipul Baghel, Jyotirmoy Amin et al.

Precise analysis of athletic motion is central to sports analytics, particularly in disciplines where nuanced biomechanical phases directly impact performance outcomes. Traditional analytics techniques rely on manual annotation or laboratory-based instrumentation, which are time-consuming, costly, and lack scalability. Automatic extraction of relevant kinetic variables requires a robust and contextually appropriate temporal segmentation. Considering the specific case of elite javelin-throw, we present a novel unsupervised framework for such a contextually aware segmentation, which applies the structured optimal transport (SOT) concept to augment the well-known Attention-based Spatio-Temporal Graph Convolutional Network (ASTGCN). This enables the identification of motion phase transitions without requiring expensive manual labeling. Extensive experiments demonstrate that our approach outperforms state-of-the-art unsupervised methods, achieving 71.02% mean average precision (mAP) and 74.61% F1-score on test data, substantially higher than competing baselines. We also release a new dataset of 211 manually annotated professional javelin-throw videos with frame-level annotations, covering key biomechanical phases: approach steps, drive, throw, and recovery.

Bikash Kumar Badatya, Vipul Baghel, Ravi Hegde

Fine-grained action localization in untrimmed sports videos presents a significant challenge due to rapid and subtle motion transitions over short durations. Existing supervised and weakly supervised solutions often rely on extensive annotated datasets and high-capacity models, making them computationally intensive and less adaptable to real-world scenarios. In this work, we introduce a lightweight and unsupervised skeleton-based action localization pipeline that leverages spatio-temporal graph neural representations. Our approach pre-trains an Attention-based Spatio-Temporal Graph Convolutional Network (ASTGCN) on a pose-sequence denoising task with blockwise partitions, enabling it to learn intrinsic motion dynamics without any manual labeling. At inference, we define a novel Action Dynamics Metric (ADM), computed directly from low-dimensional ASTGCN embeddings, which detects motion boundaries by identifying inflection points in its curvature profile. Our method achieves a mean Average Precision (mAP) of 82.66% and average localization latency of 29.09 ms on the DSV Diving dataset, matching state-of-the-art supervised performance while maintaining computational efficiency. Furthermore, it generalizes robustly to unseen, in-the-wild diving footage without retraining, demonstrating its practical applicability for lightweight, real-time action analysis systems in embedded or dynamic environments.