HyperAIHyperAI超神经

Command Palette

Search for a command to run...

算力平台

Image Captioning On Nocaps Val In Domain

评估指标

CIDEr
Pre-train (#images)
SPICE

评测结果

各个模型在此基准测试上的表现结果

Paper Title
BLIP-2 ViT-G OPT 6.7B (zero-shot)123.71.1B15.8BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image Encoders and Large Language Models
BLIP-2 ViT-G FlanT5 XL (zero-shot)123.71.1B16.3BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image Encoders and Large Language Models
BLIP-2 ViT-G OPT 2.7B (zero-shot)1231.1B15.8BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image Encoders and Large Language Models
LEMON_large116.9 200M15.8Scaling Up Vision-Language Pre-training for Image Captioning
BLIP_ViT-L114.9129M15.2BLIP: Bootstrapping Language-Image Pre-training for Unified Vision-Language Understanding and Generation
SimVLM113.71.8B-SimVLM: Simple Visual Language Model Pretraining with Weak Supervision
BLIP_CapFilt-L111.8129M14.9BLIP: Bootstrapping Language-Image Pre-training for Unified Vision-Language Understanding and Generation
LEMON_base107.7200M14.7Scaling Up Vision-Language Pre-training for Image Captioning
OmniVL104.614M15OmniVL:One Foundation Model for Image-Language and Video-Language Tasks
VinVL103.15.7M 14.2VinVL: Revisiting Visual Representations in Vision-Language Models
Enc-Dec92.615M 12.5Conceptual 12M: Pushing Web-Scale Image-Text Pre-Training To Recognize Long-Tail Visual Concepts
0 of 11 row(s) selected.
Image Captioning On Nocaps Val In Domain | SOTA | HyperAI超神经