Model card
Hybrid DLA (Shehzadi et al.).
DFKI / TU Kaiserslauternopen-sourceUnknown paramsTransformer object detector with query encoding + hybrid one-to-one/one-to-many matching
A Hybrid Approach for Document Layout Analysis in Document images. Transformer-based detection framework with enhanced contrastive learning via query encoding and hybrid training matching strategy. Achieves 97.3 mAP on PubLayNet-val — best published result. ICDAR 2024. arXiv 2404.17888.
§ 01 · Benchmarks
Every benchmark Hybrid DLA (Shehzadi et al.) has a recorded score for.
| # | Benchmark | Area · Task | Metric | Value | Rank | Date | Source |
|---|---|---|---|---|---|---|---|
| 01 | publaynet-val | Computer Vision · Document Layout Analysis | Figure | 1.0% | #1 | — | source ↗ |
| 02 | publaynet-val | Computer Vision · Document Layout Analysis | List | 1.0% | #1 | — | source ↗ |
| 03 | publaynet-val | Computer Vision · Document Layout Analysis | Overall | 1.0% | #1 | — | source ↗ |
| 04 | publaynet-val | Computer Vision · Document Layout Analysis | Table | 1.0% | #1 | — | source ↗ |
| 05 | publaynet-val | Computer Vision · Document Layout Analysis | Text | 1.0% | #1 | — | source ↗ |
| 06 | publaynet-val | Computer Vision · Document Layout Analysis | Title | 0.9% | #1 | — | source ↗ |
Rank column shows this model’s position vs all other models scored on the same benchmark + metric (competitors after the slash). #1 in red means current SOTA. Sorted by rank, then newest result.
§ 02 · Strengths by area
Where Hybrid DLA (Shehzadi et al.) actually performs.
§ 05 · Sources & freshness
Where these numbers come from.
icdar-2024
6
results
0 of 6 rows marked verified.