Sign linguistic is a language that includes the movement of hand motions. Real-time Indian Sign Linguistic recognition based on a convolutional neural network (CNN) sequential model is both interesting and valuable. The main aim is to develop a system accomplished of detecting and understanding Indian Sign Language (Ministry of Statistics & Programme Implementation, [1]) motions in real time. Indian Sign Language, like any sign language, may have a diverse set of signs. Ensuring the model is trained on a comprehensive dataset that covers various gestures is crucial. Sign language stands as a unique form of expression, not confined to verbal communication, but rather, it serves as a medium for cultural exchange and artistic representation. In this paper, we propose to address a significant need for enhancing communication for individuals with hearing impairments (Pre configure yolov5 ulralytics [2]) through the development of a real-time Indian Sign Language detection system. Ensuring comprehensive training data, real-time processing efficiency, and a user-friendly interface will be key to the success of the project (Agarwal et al. in Int J Comput Appl 116:18–22 [3]). The aim is to bridge the communication gap by creating a real-time Indian Sign Language detection system using a CNN (Sequential model) algorithm. Sign language images sourced from the web are labeled with either alphabets or numbers. The objective is to create a system that can recognize (Camgoz et al. in Proceedings of the IEEE conference on computer vision and pattern recognition (CVPR) 2018, Salt Lake City, UT, USA [4]) and interpret Indian Sign Language gestures instantly. In this paper, we firstly use CNN; in CNN, we use Modified National Institute of Standards and Technology (MNIST) database dataset train file 79.4 MB or we have test file 20.7 MB. Secondly, we have taken YOLOv5 model in which we have taken 10 images and their XML (Extensible Markup Language) file.