Research map / Vision-language and multimodal / Human and embodied modeling
Creating Images by Learning Image Semantics Using Vector Space Models
Derrall Heath, Dan Ventura
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| predict word vectors (Vector) instead of adjectives directly (Binary relevance) | visual semantic model · 145 adjectives 10-fold CV | coverage: 0.533 -> 0.628 | Hurt |
| predict word vectors (Vector) instead of adjectives directly (Binary relevance) | visual semantic model · 145 adjectives 10-fold CV | ranking loss: 0.297 -> 0.357 | Hurt |
| Vector model versus random ranking on trained adjectives | visual semantic model · 145 adjectives 10-fold CV | coverage: 0.768 -> 0.628 | Improved |
| Vector model versus random ranking on unseen adjectives | visual semantic model · Zero-shot, 10 unseen adjectives | coverage: 0.709 -> 0.444 | Improved |
| Vector model versus random ranking on unseen adjectives | visual semantic model · Zero-shot, 10 unseen adjectives | ranking loss: 0.502 -> 0.199 | Improved |
| cluster images of similar versus distinct SEEN adjectives | DARCI · SEEN adjective sets | entropy: 0.857 -> 0.714 | Mixed |
| cluster images of similar versus distinct SEEN adjectives | DARCI · SEEN adjective sets | purity: 0.360 -> 0.440 | Mixed |
| similar versus dissimilar seen partner for unseen adjectives | DARCI · UNSEEN adjectives | entropy: 0.691 -> 0.480 | Mixed |
| similar versus dissimilar seen partner for unseen adjectives | DARCI · UNSEEN adjectives | purity: 0.775 -> 0.850 | Mixed |
| similar versus dissimilar seen partner for non-adjectives | DARCI · non-adjectives | entropy: 0.828 -> 0.479 | Mixed |
| similar versus dissimilar seen partner for non-adjectives | DARCI · non-adjectives | purity: 0.680 -> 0.840 | Mixed |
Closest papers by meaning
- ConceptBed: Evaluating Concept Learning Abilities of Text-to-Image Diffusion Models (AAAI 2024 · similarity 0.87)
- A Vision Check-up for Language Models (CVPR 2024 · similarity 0.87)
- Language-Informed Visual Concept Learning (ICLR 2024 · similarity 0.86)
- Image Textualization: An Automatic Framework for Generating Rich and Detailed Image Descriptions (NeurIPS 2024 · similarity 0.86)
- Text-to-Image Generation for Abstract Concepts (AAAI 2024 · similarity 0.85)
- Learning Vision and Language Concepts for Controllable Image Generation (ICML 2025 · similarity 0.85)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
