# Dense Prediction

A depth estimation model from the Lotus model series, utilizing diffusion model technology for high-quality dense prediction

Lotus is a vision foundation model based on diffusion models, focusing on high-quality dense prediction tasks such as depth estimation. Compared to the previous version, this model is trained in disparity space (inverse depth), achieving superior performance and more stable video depth estimation.

Lotus is a vision foundation model based on diffusion models, focusing on high-quality dense prediction tasks, particularly depth estimation. Compared to the previous version, this model adopts disparity space (inverse depth) training, achieving superior performance and more stable video depth estimation.

This model employs the DPT framework with DINOv2 as the backbone network, focusing on depth estimation tasks.

A DPT model using DINOv2 as the backbone network for depth estimation tasks.

DPT framework using DINOv2 as backbone for depth estimation tasks

DPT model using DINOv2 as backbone for depth estimation tasks.

Pyramid Vision Transformer (PVT) is a vision model based on transformer architecture, specifically designed for image classification tasks.

A semantic segmentation model based on the Segformer architecture, used to assign semantic category labels to each pixel in an image.