Image Captioning On Coco Captions

评估指标

BLEU-1

BLEU-4

CIDER

METEOR

ROUGE-L

SPICE

评测结果

各个模型在此基准测试上的表现结果

							Paper Title	Repository
mPLUG	-	46.5	155.1	32.0	-	26.0	mPLUG: Effective and Efficient Vision-Language Learning by Cross-modal Skip-connections
OFA	-	44.9	154.9	32.5	-	26.6	OFA: Unifying Architectures, Tasks, and Modalities Through a Simple Sequence-to-Sequence Learning Framework
VALOR	-	-	152.5	-	-	25.7	VALOR: Vision-Audio-Language Omni-Perception Pretraining Model and Dataset
GIT	-	44.1	151.1	32.2	-	26.3	GIT: A Generative Image-to-text Transformer for Vision and Language
VAST	-	-	149.0	-	-	27.0	VAST: A Vision-Audio-Subtitle-Text Omni-Modality Foundation Model and Dataset
BLIP-2 ViT-G OPT 2.7B (zero-shot)	-	43.7	145.8	-	-	-	BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image Encoders and Large Language Models
LEMON	-	42.6	145.5	31.4	-	25.5	Scaling Up Vision-Language Pre-training for Image Captioning	-
BLIP-2 ViT-G OPT 6.7B (zero-shot)	-	43.5	145.2	-	-	-	BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image Encoders and Large Language Models
BLIP-2 ViT-G FlanT5 XL (zero-shot)	-	42.4	144.5	-	-	-	BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image Encoders and Large Language Models
GRIT (No VL pretraining - base)	84.2	42.4	144.2	30.6	60.7	24.3	GRIT: Faster and Better Image captioning Transformer Using Dual Visual Features
ExpansionNet v2 (No VL pretraining)	83.5	42.7	143.7	30.6	61.1	24.7	Exploiting Multiple Sequence Lengths in Fast End to End Training for Image Captioning
CoCa	-	40.9	143.6	33.9	-	24.7	CoCa: Contrastive Captioners are Image-Text Foundation Models
SimVLM	-	40.6	143.3	33.4	-	25.4	SimVLM: Simple Visual Language Model Pretraining with Weak Supervision
Xmodal-Ctx + OSCAR	-	41.3	142.2	-	-	24.9	Beyond a Pre-Trained Object Detector: Cross-Modal Textual and Visual Context for Image Captioning
Prompt Tuning	-	41.81	141.4	31.51	-	24.42	Prompt Tuning for Generative Multimodal Pretrained Models
VinVL	-	41.0	140.9	31.1	-	25.2	VinVL: Revisiting Visual Representations in Vision-Language Models
X-VLM (base)	-	41.3	140.8	-	-	-	Multi-Grained Vision Language Pre-Training: Aligning Texts with Visual Concepts
Oscar	-	41.7	140	30.6	-	24.5	Oscar: Object-Semantics Aligned Pre-training for Vision-Language Tasks
Xmodal-Ctx	83.4	41.4	139.9	30.4	60.4	24.0	Beyond a Pre-Trained Object Detector: Cross-Modal Textual and Visual Context for Image Captioning
Prismer	-	40.4	136.5	31.4	-	24.4	Prismer: A Vision-Language Model with Multi-Task Experts

0 of 40 row(s) selected.