Image Captioning Web Application using Deep Learning Algorithms
Surya R.E., S. B. Mahalakshmi · International Journal of Computer Applications · 2023
Images and visual signs play a vital role in communication and comprehension, but they pose challenges for visually impaired individuals.This paper presents an innovative solution, leveraging Convolutional Neural Networks (CNN) and Long Short-Term Memory (LSTM) networks, to create a photo-tospeech application that enhances the quality of life for individuals with visual impairments.The development, methodology, and evaluation of this application, demonstrates its potential to provide real-time image captions and improve accessibility.This work includes Flickr 8k dataset for training the model (VGG16) and attains a BLEU score of 56%. General TermsThe paper discusses an innovative solution for visually impaired individuals, using Pattern Recognition, Security, and Algorithms.It leverages CNN and LSTM networks to create a photo-to-speech application to improve accessibility.