HyperAI超神経
ホームニュース最新論文チュートリアルデータセット百科事典SOTALLMモデルGPU ランキング学会
検索
サイトについて
HyperAI超神経

Image Captioning On Nocaps Val Near Domain

評価指標

CIDEr
Pre-train (#images)
SPICE

評価結果

このベンチマークにおける各モデルのパフォーマンス結果

モデル名
CIDEr
Pre-train (#images)
SPICE
Paper TitleRepository
BLIP-2 ViT-G OPT 6.7B (zero-shot)119.21.1B15.3BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image Encoders and Large Language Models
OmniVL108.314M14.9OmniVL:One Foundation Model for Image-Language and Video-Language Tasks-
VinVL96.15.7M13.8VinVL: Revisiting Visual Representations in Vision-Language Models
BLIP-2 ViT-G FlanT5 XL (zero-shot)120.21.1B15.9BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image Encoders and Large Language Models
Enc-Dec88.3-12.1Conceptual 12M: Pushing Web-Scale Image-Text Pre-Training To Recognize Long-Tail Visual Concepts
BLIP_ViT-L112.1129M 14.9BLIP: Bootstrapping Language-Image Pre-training for Unified Vision-Language Understanding and Generation
LEMON_large113.3200M 15.1Scaling Up Vision-Language Pre-training for Image Captioning-
SimVLM110.91.8B-SimVLM: Simple Visual Language Model Pretraining with Weak Supervision
BLIP_CapFilt-L108.6129M14.8BLIP: Bootstrapping Language-Image Pre-training for Unified Vision-Language Understanding and Generation
BLIP-2 ViT-G OPT 2.7B (zero-shot)117.81.1B15.4BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image Encoders and Large Language Models
0 of 10 row(s) selected.