Alloy Das (অলয়)

Mammalian hair serves as a critical biological marker, aiding species identification essential for wildlife conservation and crime control. This study introduces the first extensive benchmark for classifying microscopic images of mammal hair from species prioritized for conservation. Our goal is to develop standardized methods, metrics, and best practices for utilizing advanced computer vision techniques, including Convolutional Neural Networks (CNNs), Vision Transformers (ViTs) , and Swin Transformers, to classify hair samples across Order, Family, Genus and Species taxonomic levels. We present a novel dataset of 76 species, including critically endangered and endangered species, curated specifically for this classification challenge. The methodology integrates automated feature extraction of cuticle patterns and medulla structures, enabling high-precision species differentiation. Our findings demonstrate that Swin Transformer-based models outperform traditional CNNs and ViTs across taxonomic levels, with techniques like image cropping further improving classification accuracy by diversifying the training set. The proposed Tricho-Vision framework offers significant applications in biodiversity monitoring and wildlife crime investigation, facilitating accurate species identification from forensic hair samples. Additionally, we introduce a interactive tool for real-time taxonomic classification, showcasing the practical utility of our research and fostering broader interdisciplinary engagement in conservation science and forensic applications. • Curated dataset with 76 species for research in hair classification. • Standardized suite for evaluating Tricho-Taxonomy models. • Exhaustive tests ensure framework performance accuracy. • Real-time demo highlights practical conservation applications.

Scene Text Editing (STE) is a challenging research prob-lem, that primarily aims towards modifying existing texts in an image while preserving the background and the font style of the original text. Despite its utility in numerous real-world applications, existing style-transfer-based approaches have shown sub-par editing performance due to (1) complex image backgrounds, (2) diverse font attributes, and (3) varying word lengths within the text. To address such limitations, in this paper, we propose a novel font-agnostic scene text editing and rendering framework, named FASTER, for simultaneously generating text in arbitrary styles and locations while preserving a natural and realistic appearance and structure. A combined fusion of target mask generation and style transfer units, with a cascaded self-attention mech-anism has been proposed to focus on multi-level text region edits to handle varying word lengths. Extensive evaluation on a real-world database withfurther subjective human eval-uation study indicates the superiority of FASTER in both scene text editing and rendering tasks, in terms of model per-formance and efficiency. The code and pre-trained models have been released in our Gi thub repo.

When used in a real-world noisy environment, the capacity to generalize to multiple domains is essential for any autonomous scene text spotting system. However, existing state-of-the-art methods employ pretraining and fine-tuning strategies on natural scene datasets, which do not exploit the feature interaction across other complex domains. In this work, we explore and investigate the problem of domain-agnostic scene text spotting, i.e., training a model on multi-domain source data such that it can directly generalize to target domains rather than being specialized for a specific domain or scenario. In this regard, we present the community a text spotting validation benchmark called Under-Water Text (UWT) for noisy underwater scenes to establish an important case study. Moreover, we also design an efficient super-resolution based end-to-end transformer baseline called DA-TextSpotter which achieves comparable or superior performance over existing text spotting architectures for both regular and arbitrary-shaped scene text spotting benchmarks in terms of both accuracy and model efficiency. The dataset, code and pre-trained models have been released in our Github.

The adaptation capability to a wide range of domains is crucial for scene text spotting models when deployed to real-world conditions. However, existing SOTA approaches usually incorporate scene text detection and recognition simply by pretraining on natural scene text datasets, which do not directly exploit the intermediate feature representations between multiple domains. Here, we investigate the problem of domain-adaptive scene text spotting, i.e., training a model on multi-domain source data such that it can directly adapt to target domains rather than being specialized for a specific domain or scenario. Further, we investigate a transformer baseline called Swin-TESTR to focus on solving scene-text spotting for both regular and arbitraryshaped text along with an exhaustive evaluation. The results demonstrate the potential of intermediate representations to gain significant performance on text spotting benchmarks across multiple domains (e.g. language, synth-to-real, and documents). both in terms of accuracy and efficiency.