Overview#
TAO is a pre-training, finetuning, and optimization application for computer vision DNNs (deep neural networks). The finetuning pipelines in TAO are implemented with the PyTorch Deep Learning Framework. The details of these pipelines, including hyperparameters and features, are covered in the subsequent sections.
The source code for these networks are hosted on GitHub.
Pretrain backbones from unlabeled data for downstream tasks.
Classify images into categories using transformer and CNN backbones.
Detect and localize objects in images with bounding boxes.
Semantic and instance segmentation of scenes and objects.
Detect changes between image pairs for classification and segmentation.
Estimate per-pixel depth and disparity from images.
Detect, reconstruct, and estimate pose from 3D and LiDAR data.
Recognize and re-identify objects, people, characters, and poses.
Detect defects and anomalies for automated optical inspection.
Pretrain backbones from unlabeled data for downstream tasks.