Research map / Point clouds / Point cloud classification
A Machine Learning Paradigm for Studying Pictorial Realism: How Accurate are Constable's Clouds?
Zhuomin Zhang, Elizabeth Mansfield, Jia Li, John Russell, George S. Young, Catherine Adams, Kevin A. Bowley, James Z. Wang
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 |
|---|---|---|---|
| add sky selection (ablation) | SFF-CNN · CCSN cloud photos (20% test) | precision: 0.938 -> 0.972 | Improved |
| add semi-supervised learning with unlabeled cloud photos (ablation) | SFF-CNN · CCSN cloud photos (20% test) | precision: 0.944 -> 0.972 | Improved |
| add edge feature fusion (ablation) | SFF-CNN · CCSN cloud photos (20% test) | precision: 0.955 -> 0.972 | Improved |
| add edge feature fusion (ablation) | SFF-CNN · CCSN cloud photos (20% test) | recall: 0.953 -> 0.969 | Improved |
| use SFF-CNN instead of CloudNet | CCSN classifier · CCSN cloud photos (20% test) | precision: 0.891 -> 0.972 | Improved |
| use SFF-CNN instead of ensemble learning classifier | CCSN classifier · CCSN cloud photos (20% test) | precision: 0.953 -> 0.972 | Improved |
| compare Constable with Boudin on classification accuracy | painting cloud classification · Painting cloud types | accuracy: p = 0.024 | Improved |
| compare Constable with Valenciennes on classification accuracy | painting cloud classification · Painting cloud types | accuracy: p = 0.332 | No clear effect |
| add edge features for Constable's paintings | SFF-CNN · Painting cloud types | accuracy: p = 0.122 | Improved |
| compare Constable with Valenciennes on style distance to cloud photos | Rstyle · Paintings Rstyle | Rstyle: 1.163 -> 1.191 | Hurt |
| compare Constable with Cox on style distance to cloud photos | Rstyle · Paintings Rstyle | Rstyle: 1.319 -> 1.191 | Improved |
| compare Lionel Constable with Lee on style distance to John Constable | Dstyle · Style distance Dstyle | Dstyle: 0.439 -> 0.359 | Improved |
Closest papers by meaning
- The Photographer's Eye: Teaching Multimodal Large Language Models to See, and Critique Like Photographers (CVPR 2025 · similarity 0.83)
- PoSh: Using Scene Graphs to Guide LLMs-as-a-Judge for Detailed Image Descriptions (ICLR 2026 · similarity 0.81)
- APDDv2: Aesthetics of Paintings and Drawings Dataset with Artist Labeled Scores and Comments (NeurIPS 2024 · similarity 0.81)
- Shadows Don't Lie and Lines Can't Bend! Generative Models don't know Projective Geometry...for now (CVPR 2024 · similarity 0.80)
- Semi-Truths: A Large-Scale Dataset of AI-Augmented Images for Evaluating Robustness of AI-Generated Image detectors (NeurIPS 2024 · similarity 0.80)
- Position: What makes an image realistic? (ICML 2024 · similarity 0.80)
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.
