Intel/zoedepth-nyu-kitti is a depth estimation model hosted on Hugging Face. It addresses the task of predicting depth maps from single RGB images, a core problem in computer vision.
The model is a version of ZoeDepth fine-tuned on the NYU and KITTI datasets. It belongs to the class of depth estimation models and is implemented using a Transformer-based architecture. Users can access it through the Transformers library via a high-level pipeline for depth estimation or by loading the processor and model classes directly. Code examples show initialization with AutoImageProcessor and AutoModelForDepthEstimation, including support for automatic device mapping.
It is provided as an open model card with associated files in Safetensors format. The license is MIT. The page lists integration options including notebooks on Google Colab and Kaggle. A related arXiv paper is cited with identifier 2302.12288.
The model is intended for research and development in depth estimation tasks within the Transformers ecosystem.
In the Other AI space, Zoedepth Nyu Kitti takes a focused approach. It focuses on estimating depth from a single image without specialized hardware. Zoedepth Nyu Kitti is an open-source project aimed at developers. The project is open source (MIT). It runs on the web and API.
It is developed by Intel (United States), and the product first shipped in 2022. The GitHub repository has been archived. Among its 3 catalogued features are Depth Estimation, Monocular Depth, and transformers.
Latest indexed changes and source events
Intel/zoedepth-nyu-kitti verified by the PulseGate indexer
⚠ Archived
Other apps tracked under the same category.
