AI Paper Atlas

Research map / Benchmarks / CORD

CORD leaderboard

10 papers at ICML, NeurIPS, ICLR, CVPR, AAAI, ICCV and ECCV, and the journals TPAMI, IJCV and JMLR report results on CORD. Best F1: 98.75 (UDoc: MSM + VCL + VLA, NeurIPS 2021).

Numbers are as each paper reports them, read by AI from its results tables and not checked by hand. Papers can differ in setting (training data, model size, extra data); the topic column shows what each paper is about. Open the paper before you compare or quote a number.

CORD · F1

10 papers, each with its best reported number for its own method; higher is better.

#F1 ↑MethodPaperTopic
198.75UDoc: MSM + VCL + VLAUniDoc: Unified Pretraining Framework for Document Understanding
NeurIPS 2021 · Table: Experimental results an
Human and embodied modeling
297.7DocFormerv2-largeDocFormerv2: Local Features for Document Understanding
AAAI 2024 · Table: CORD dataset comparison
Human and embodied modeling
397.58UDOP
T5-large
Unifying Vision, Text, and Layout for Universal Document Processing
CVPR 2023 · Table: Performance on FUNSD, C
Human and embodied modeling
497.04AETNet_base
LayoutLMv3_base
Alignment-Enriched Tuning for Patch-Level Pre-trained Document Image Models
AAAI 2023 · Table: Performance comparison
CLIP and zero-shot
597DocMambaDocMamba: Efficient Document Pre-training with State Space Model
AAAI 2025 · Table 1
Attention and transformers
696.41SAIL
GPT-4
SAIL: Sample-Centric In-Context Learning for Document Information Extraction
AAAI 2025 · Quantitative results table
Language, safety and interpretability
795.8LDM
SAM-BASE
LDP: Generalizing to Multilingual Visual Information Extraction by Language Decoupled Pretraining
AAAI 2025 · Table, Performance on CORD
Human and embodied modeling
889.3HRVDA
LLaMA-2-7B
HRVDA: High-Resolution Visual Document Assistant
CVPR 2024 · Table: Comparison of HRVDA wit
Multimodal LLMs
984.9SeRum-prompt
Swin-B + mBART
Attention Where It Matters: Rethinking Visual Document Understanding with Selective Region Concentration
ICCV 2023 · Table 1
Human and embodied modeling
1084.47Qwen2-VL-7B+RIDGE
Qwen2-VL-7B
Relation-Rich Visual Document Generator for Visual Information Extraction
CVPR 2025 · Table 2
Human and embodied modeling

Not affiliated with any conference, journal or the benchmark's authors. Corrections: contact@latenticlabs.com.