ML Research Wiki / Benchmarks / Image Captioning / COCO (Common Objects in Context)

COCO (Common Objects in Context)

Image Captioning Benchmark

Performance Over Time

📊 Showing 16 results | 📏 Metric: CIDEr

Top Performing Models

Rank Model Paper CIDEr Date Code
1 ExpansionNet v2 Exploiting Multiple Sequence Lengths in Fast End to End Training for Image Captioning 143.70 2022-08-13 📦 jchenghu/expansionnet_v2
2 M2 Transformer Meshed-Memory Transformer for Image Captioning 131.20 2019-12-17 📦 aimagelab/meshed-memory-transformer 📦 Japanese-Image-Captioning/M2-transformer-for-Japanese
3 RDN Reflective Decoding Network for Image Captioning 125.20 2019-08-30 -
4 Lyrics Lyrics: Boosting Fine-grained Language-Vision Alignment and Comprehension via Semantic-aware Visual Objects 121.10 2023-12-08 -
5 Flamingo (80B; 4-shot) Retrieval-Augmented Multimodal Language Modeling 103.00 2022-11-22 -
6 RA-CM3 (2.7B) Retrieval-Augmented Multimodal Language Modeling 89.10 2022-11-22 -
7 Flamingo (3B; 4-shot) Retrieval-Augmented Multimodal Language Modeling 85.00 2022-11-22 -
8 Parti Retrieval-Augmented Multimodal Language Modeling 83.90 2022-11-22 -
9 Vanilla CM3 Retrieval-Augmented Multimodal Language Modeling 71.90 2022-11-22 -
10 X-LXMERT Retrieval-Augmented Multimodal Language Modeling 55.80 2022-11-22 -

All Papers (16)

jchenghu/expansionnet_v2
aimagelab/meshed-memory-transformer Japanese-Image-Captioning/M2-transformer-for-Japanese
PaddlePaddle/PaddleNLP PaddlePaddle/Research PaddlePaddle/Research
lucidrains/imagen-pytorch google-research/pix2seq
rwightman/pytorch-image-models pytorch/vision