Enhancing Image Caption Generation Using Reinforcement Learning with Human Feedback
This work addresses the challenge of producing human-preferred captions for images, which is an incremental improvement in the domain of human-aligned generative AI models.
The researchers tackled the problem of generating image captions that align with human preferences by integrating supervised learning and reinforcement learning with human feedback (RLHF) on the Flickr8k dataset, introducing a novel loss function to optimize the model based on human feedback.
Research on generative models to produce human-aligned / human-preferred outputs has seen significant recent contributions. Between text and image-generative models, we narrowed our focus to text-based generative models, particularly to produce captions for images that align with human preferences. In this research, we explored a potential method to amplify the performance of the Deep Neural Network Model to generate captions that are preferred by humans. This was achieved by integrating Supervised Learning and Reinforcement Learning with Human Feedback (RLHF) using the Flickr8k dataset. Also, a novel loss function that is capable of optimizing the model based on human feedback is introduced. In this paper, we provide a concise sketch of our approach and results, hoping to contribute to the ongoing advances in the field of human-aligned generative AI models.