• Amitesh Deshmukh
• Aaditya Jadhav
• Anagha Chaudhari
• Sujit Deore
• Uddhav Patil

: In this study, we have developed an Image Captioning model leveraging InceptionV3 for image classification and extracting relevant features. The model incorporates Recurrent Neural Networks (RNNs), specifically Long Short-Term Memory (LSTM) networks, to generate descriptive captions in English for the images processed. Additionally, we have employed GloVe word embeddings to enhance the linguistic representation within the model. The system classifies images by identifying elements within them and generates meaningful statements to describe the fed images. The generated captions hold potential applications in various domains, such as social media posts, blog articles, accessibility for visually impaired individuals, e-commerce, advertising, surveillance, medical diagnosis, education, among others. The model is trained using Categorical Cross Entropy Loss, aiming to optimize the accuracy of caption predictions. As the model evolves, it is anticipated that it will provide more accurate descriptions for images of diverse formats and content, contributing to its versatility and utility across a wide array of applications

Image Captioning, InceptionV3, Recurrent Neural Networks (RNNs), Long Short-Term Memory (LSTM), GloVe, Categorical Cross Entropy Loss, Gated Recurrent Unit(GRU), Encoder, Decoder, Image captioning.

"SpeechSnap- Describing Images with voice and multi-language support features", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.11, Issue 5, page no.a798-a811, May-2024, Available :http://www.jetir.org/papers/JETIR2405095.pdf

"SpeechSnap- Describing Images with voice and multi-language support features", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.11, Issue 5, page no. ppa798-a811, May-2024, Available at : http://www.jetir.org/papers/JETIR2405095.pdf