[ECCV'20] Patch-match and Plane-regularization for Unsupervised Indoor Depth Estimation
SharpNet: Fast and Accurate Recovery of Occluding Contours in Monocular Depth Estimation
Joint scene classification and semantic segmentation with FuseNet
nyuv2 toolbox for data extraction and loading.
PyTorch Implementation of "NDDR-CNN: Layerwise Feature Fusing in Multi-Task CNNs by Neural Discriminative Dimensionality Reduction"
The NyuV2 Raw Dataset Extractor is a Python project that processes the raw dataset from NyuV2. It provides operations like depth map colorization, undistortion, and frame synchronization to create video sequences. This Python adaptation of the official NyuV2 toolbox simplifies dataset manipulation.
Replicated results from DenseDepth using DenseNet169 in Python.
MTLA - Multi-Task Learning Archive
Staged fine-tuning of YOLO26-Depth and Depth Anything V2 for indoor monocular metric depth estimation — NYUv2 → ScanNet → real lab/office environments, for robotics deployment.
To associate your repository with the nyuv2 topic, visit your repo's landing page and select "manage topics."