The paper presents research on “Hand Sign Detection for Deaf and Dumb,” using deep learning techniques, which make communication easier between “Deaf and Dumb” community and general audience [normal people]. We implement this by using Convolutional Neural networks [CNN] and Recurrent Neural Networks [RNN]. We use American Sign Language [ASL] to generate hand gestures. We train our model using American Sign language data set, implemented using “Transfer learning” techniques. Our model showcases hand gestures as output which are understandable by deaf and dumb people. It leverages the gap or acts as a communication medium between general audience and deaf and dumb people. Accuracy, precision, recall indicate the system efficiency in translating ASL gestures in real-world scenarios. The paper discussed mainly focuses on assistive techniques to deaf and dumb people. Considering the challenges faced by individuals with hearing impairments, this model assists them in their regular activities. By utilizing deep learning techniques, the way sign language translator offers an assured solution for global communication. This breaks the bridge between normal people and targeted audience [deaf and dumb people]. This paper brings a platform for better communication for individuals with hearing impairments.