Research map / Vision-language and multimodal / Human and embodied modeling
Ground-Truth or DAER: Selective Re-Query of Secondary Information
Stephan J. Lemmer, Jason J. Corso
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 the regression loss to the correctness loss (ablation) | DAER · KCVE (PASCAL3D+ val, crowdsourced keypoints) | AMAE: 0.2937 -> 0.2864 | Improved |
| add the regression loss to the correctness loss (ablation) | DAER · HSC (SUN397) | AMAE: 1.79 -> 1.62 | Improved |
| add the correctness loss (dual loss) to direct regression (ablation) | DAER · KCVE (PASCAL3D+ val, crowdsourced keypoints) | AMAE: 1.1633 -> 0.2864 | Improved |
| add the correctness loss (dual loss) to direct regression (ablation) | DAER · HSC (SUN397) | AMAE: 2.05 -> 1.62 | Improved |
| give the rejection model access to the seed (ablation) | DAER · KCVE (PASCAL3D+ val, crowdsourced keypoints) | AMAE: 0.8002 -> 0.2864 | Improved |
| give the rejection model access to the seed (ablation) | DAER · HSC (SUN397) | AMAE: 2.28 -> 1.62 | Improved |
| use DAER instead of the next-best baseline (Coarse Entropy) at target MAE 5 | DAER · HSC (SUN397) | percentage of seeds at target MAE: 6.7% -> 5.1% | Improved |
| use DAER instead of Task Network Percentile | DAER · KCVE (PASCAL3D+ val, crowdsourced keypoints) | AMAE: 0.3092 -> 0.2864 | Improved |
Closest papers by meaning
- UniBench: Visual Reasoning Requires Rethinking Vision-Language Beyond Scaling (NeurIPS 2024 · similarity 0.82)
- Focusing Where Vision Matters: Selective Training for Large Vision Language Models via Visual Information Gain (ICML 2026 · similarity 0.82)
- Reflexive Guidance: Improving OoDD in Vision-Language Models via Self-Guided Image-Adaptive Concept Generation (ICLR 2025 · similarity 0.82)
- Generating Images of Rare Concepts Using Pre-trained Diffusion Models (AAAI 2024 · similarity 0.82)
- AVA-Bench: Atomic Visual Ability Benchmark for Vision Foundation Models (CVPR 2026 · similarity 0.82)
- Rex-Thinker: Grounded Object Referring via Chain-of-Thought Reasoning (ICLR 2026 · similarity 0.82)
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.
