Taegyeong Lee

Taegyeong Lee, Soyeong Kwon, Taehwan Kim

Recent advances in the diffusion models have significantly improved text-to-image generation. However, generating videos from text is a more challenging task than generating images from text, due to the much larger dataset and higher computational cost required. Most existing video generation methods use either a 3D U-Net architecture that considers the temporal dimension or autoregressive generation. These methods require large datasets and are limited in terms of computational costs compared to text-to-image generation. To tackle these challenges, we propose a simple but effective novel grid diffusion for text-to-video generation without temporal dimension in architecture and a large text-video paired dataset. We can generate a high-quality video using a fixed amount of GPU memory regardless of the number of frames by representing the video as a grid image. Additionally, since our method reduces the dimensions of the video to the dimensions of the image, various image-based methods can be applied to videos, such as text-guided video manipulation from image manipulation. Our proposed method outperforms the existing methods in both quantitative and qualitative evaluations, demonstrating the suitability of our model for real-world video generation.

Taegyeong Lee, Jiwon Park, Kyunga Bang et al.

Recent advances in generative models have enabled significant progress in tasks such as generating and editing images from text, as well as creating videos from text prompts, and these methods are being applied across various fields. However, in the financial domain, there may still be a reliance on time-series data and a continued focus on transformer models, rather than on diverse applications of generative models. In this paper, we propose a novel approach that leverages text-to-image model by treating time-series data as a single image pattern, thereby enabling the prediction of stock price trends. Unlike prior methods that focus on learning and classifying chart patterns using architectures such as ResNet or ViT, we experiment with generating the next chart image from the current chart image and an instruction prompt using diffusion models. Furthermore, we introduce a simple method for evaluating the generated chart image against ground truth image. We highlight the potential of leveraging text-to-image generative models in the financial domain, and our findings motivate further research to address the current limitations and expand their applicability.