Two-step image grounding pipeline: extracts referring expressions from (image, caption) pairs and grounds them to pixel-space bounding boxes via a VLM. Use when the user wants to ground captions to bboxes, generate phrase-grounded annotations, auto-label images for grounding, or run the image_grounding pipeline.
Install
npx skills add NVIDIA/skills --skill tao-generate-image-groundingnpx skills add NVIDIA/skills --skill tao-generate-image-grounding -a claude-code -gnpx skills add NVIDIA/skills --skill tao-generate-image-grounding -a codex -ggemini skills install https://github.com/NVIDIA/skills.git --path skills/tao-generate-image-groundinggh skill install NVIDIA/skills tao-generate-image-groundinggit clone --depth 1 https://github.com/NVIDIA/skills /tmp/NVIDIA__skills
mkdir -p ~/.claude/skills ~/.agents/skills
cp -R /tmp/NVIDIA__skills/skills/tao-generate-image-grounding ~/.claude/skills/tao-generate-image-grounding # Claude Code
cp -R /tmp/NVIDIA__skills/skills/tao-generate-image-grounding ~/.agents/skills/tao-generate-image-grounding # Codex, Cursor, Gemini CLI, Copilot, OpenCode