
Deepstream dev
SkillGuides an agent through building NVIDIA DeepStream video analytics pipelines in Python
- Open source
- 4K GitHub stars
- Computer vision


Guides an agent through building NVIDIA DeepStream video analytics pipelines in Python

Runs NVIDIA's DeepStream Multi-View 3D Tracking app on sample or custom multi-camera video

Trains and runs NVIDIA TAO Mask Grounding DINO for text-prompted instance segmentation

Trains, evaluates and runs Co-DETR object detectors in NVIDIA TAO containers

Runs NVIDIA TAO Grounding DINO detection through an iterative data-mining and retraining

Trains, evaluates, exports and runs ST-GCN models that classify skeleton pose sequences

Trains vision transformers with self-supervised NVDINOv2 learning on unlabeled images

Trains, evaluates and exports OCRNet scene text recognition models with TAO

Trains, evaluates, exports and runs person re-identification models in Docker

Trains, evaluates and exports TAO monocular depth models from single RGB images

Trains, evaluates and exports Mask2Former segmentation models through NVIDIA TAO

Trains, evaluates, exports and runs TAO models that classify actions in video clips
The closest alternatives to Deepstream import vision model are Deepstream dev, Deepstream run mv3dt, Tao train mask grounding dino, Tao train codetr and Tao run deft object detection.
Yes. Deepstream dev, Deepstream run mv3dt, Tao train mask grounding dino have a free or open-source option.