MiDaS
A foundation model for monocular depth estimation developed by Intel Labs and collaborators. It infers relative depth information from a single image.
Timeline (last 90 days)
Open-source models such as Depth Anything and MiDaS have significantly lowered the barrier to using monocular depth estimation.
Representative open-source models for monocular depth estimation include MiDaS and Depth Anything, which can run locally on consumer-grade GPUs.
All Facts (4)
深度重绘使用MiDaS、DPT等单目深度估计模型从原图中自动推断深度图
85%UnverifiedModern AI video generation models use variational autoencoders to compress high-dimensional pixel images into low-dimensional latent vector representations, finding transition paths in the latent space rather than performing pixel-level linear interpolation.
60%UnverifiedOpen-source models such as Depth Anything and MiDaS have significantly lowered the barrier to using monocular depth estimation.
50%UnverifiedRepresentative open-source models for monocular depth estimation include MiDaS and Depth Anything, which can run locally on consumer-grade GPUs.
50%