Research map / Probabilistic and scientific modeling / Molecules and proteins
Deep Structured Prediction with Nonlinear Output Transformations
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 explicit pairwise structure to the Unary model | Word recognition · Synthetic word recognition test, chain | word accuracy: 0.000 -> 0.040 | Helped |
| Replace the linear top transformation with a nonlinear one | Structured word recognition model, chain · Synthetic word recognition test | word accuracy: 0.085 -> 0.075 | Hurt |
| Replace the linear top transformation with a nonlinear one | Structured word recognition model, chain · Synthetic word recognition test | character accuracy: 0.4030 -> 0.4150 | Helped |
| Replace the linear top transformation with a nonlinear one | Structured word recognition model, second-order · Synthetic word recognition test | word accuracy: 0.090 -> 0.140 | Helped |
| Replace the linear top transformation with a nonlinear one | Structured word recognition model, second-order · Synthetic word recognition test | character accuracy: 0.4090 -> 0.4420 | Helped |
| Add a global potential transformation T | Image tagging · MIRFLICKR25k | Hamming loss | Helped |
| Use SPEN-like relaxed gradient-ascent inference instead of the proposed inference | Image tagging NLTop · MIRFLICKR25k | Hamming loss | Hurt |
| Increase DeepStruct parameters with two additional potential-transforming MLPs | Image tagging · MIRFLICKR25k | Hamming loss: 2,444,544 vs 1,329,408 | Hurt |
| Add the potential transformation T | Semantic segmentation · Weizmann Horses | mean intersection-over-union | Helped |
| Use SPEN-like relaxed gradient-ascent inference instead of the proposed inference | NLStruct · Weizmann Horses | mean intersection-over-union | No clear effect |
Closest papers by meaning
- Adversarial Localized Energy Network for Structured Prediction (AAAI 2020 · similarity 0.88)
- Graph Structured Prediction Energy Networks (NeurIPS 2019 · similarity 0.88)
- Non-parametric Structured Output Networks (NeurIPS 2017 · similarity 0.88)
- Structured Prediction Energy Networks (ICML 2016 · similarity 0.87)
- Structured Convolutions for Efficient Neural Network Design (NeurIPS 2020 · similarity 0.87)
- SOInter: A Novel Deep Energy-Based Interpretation Method for Explaining Structured Output Models (ICLR 2024 · similarity 0.86)
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.