Closed wududu123 closed 4 months ago
The author says in the paper: "Compared with the latest models [31] built on Stable Diffusion, our models are significantly more efficient (more than 10× faster) and more accurate." It is 10x faster than Stable Diffusion based models, not 10x faster than depth anything v1.
I saw some news on Toutiao is that TikTok has launched version depthanthing v2, which is 10 times faster.
So That's all fake news.???