Hand gesture classification using a novel CNN-crow search algorithm

Abstract Human–computer interaction (HCI) and related technologies focus on the implementation of interactive computational systems. The studies in HCI emphasize system use, creation new techniques that support user activities, access to information, ensures seamless communication. use artificial intelligence deep learning-based models has been extensive across various domains yielding state-of-the-art results. In present study, a crow search-based convolution neural networks model implemented gesture recognition pertaining domain. hand dataset used study is publicly available one, downloaded from Kaggle. this work, one-hot encoding technique convert categorical data values binary form. This followed by search algorithm (CSA) for selecting optimal hyper-parameters training using networks. irrelevant parameters are eliminated consideration, which contributes towards enhancement accuracy classifying gestures. generates 100 percent testing justifies superiority against traditional models.