HyperAI超神经

首页算力平台文档资讯论文教程数据集百科 SOTA LLM 模型天梯 GPU 天梯顶会

中文

HyperAI超神经

Image Retrieval On Cirr

评估指标

(Recall@5+Recall_subset@1)/2

评测结果

各个模型在此基准测试上的表现结果

		Paper Title	Repository
SPN4CIR	82.69	Improving Composed Image Retrieval via Contrastive Learning with Scaling Positives and Negatives
SPN4CIR (SPRC)	82.69	Improving Composed Image Retrieval via Contrastive Learning with Scaling Positives and Negatives
SPRC2	82.66	Sentence-level Prompts Benefit Composed Image Retrieval
SPRC	81.39	Sentence-level Prompts Benefit Composed Image Retrieval
Candidate Set Re-ranking	80.9	Candidate Set Re-ranking for Composed Image Retrieval with Dual Multi-modal Encoder
CaLa	78.74	CaLa: Complementary Association Learning for Augmenting Composed Image Retrieval
CASE (Pre-trained on LaSCo.Ca)	78.25	Data Roaming and Quality Assessment for Composed Image Retrieval
CASE	77.5	Data Roaming and Quality Assessment for Composed Image Retrieval
VISTA (base)	75.9	VISTA: Visualized Text Embedding For Universal Multi-Modal Retrieval
TG-CIR (Wen et al., 2023)	75.6	Target-Guided Composed Image Retrieval	-
CLIP4Cir (v3)	75.10	Composed Image Retrieval using Contrastive Learning and Task-oriented CLIP-based Features
BLIP4CIR+Bi	72.59	Bi-directional Training for Composed Image Retrieval via Text Prompt Learning
CLIP4Cir (v2)	69.09	Conditioned and Composed Image Retrieval Combining and Partially Fine-Tuning CLIP-Based Features	-
CLIP4Cir	63.87	Effective Conditioned and Composed Image Retrieval Combining CLIP-Based Features	-
CIRPLANT	45.88	Image Retrieval on Real-life Images with Pre-trained Vision-and-Language Models
ARTEMIS	43.05	ARTEMIS: Attention-based Retrieval with Text-Explicit Matching and Implicit Similarity
MMRet-MLLM	-	MegaPairs: Massive Data Synthesis For Universal Multimodal Retrieval

0 of 17 row(s) selected.