I started getting into image generation last year with Stable Diffusion's 1.5 release last October, but seeing the improvements from Midjourney and DallE 3 is extremely impressive. The image to video models from Runway and Pika labs & Stable Video Diffusion are reminding me of Midjourney prior to V4, hoping to see similar improvements in 2024 for those models.
2 comments
[ 3.1 ms ] story [ 16.0 ms ] thread