Audio and Caption Generation using ML
R P Rajeshwari, Thammineni Pushyamitra, G. M. G. Madhuri, B Pavani Kumari · International Journal of Advanced Research in Science Communication and Technology · 2022
Using deep learning and computer vision, image caption generation identifies the context of an image and comments on it with the most relevant captions. It entails an image's epithet combined with English keywords, as well as datasets provided during model training. We produce a vectorized representation of an image using a deep Convolutional neural network, which is then fed into a long-short-term memory (LSTM) network, which provides captions. LSTM (Long short-term memory) is also one of the RNN (Recurrent Neural Network) constructs, mainly utilized in the deep learning discipline. The vivid applications of image captioning are natural language processing applications, recommendations in editing applications, helpful for virtual assistants, for indexing the image, for the blind or partially-sighted people, for the internet community, etc.