Zero Shot Composed Image Retrieval Zs Cir On

Metriken

mAP@10

Ergebnisse

Leistungsergebnisse verschiedener Modelle zu diesem Benchmark

Modellname	mAP@10	Paper Title	Repository
iSEARLE-OTI (CLIP B/32)	10.94	iSEARLE: Improving Textual Inversion for Zero-Shot Composed Image Retrieval
SEIZE (CLIP G/14 & GPT-4o)	37.23	Semantic Editing Increment Benefits Zero-Shot Composed Image Retrieval
iSEARLE (CLIP B/32)	11.24	iSEARLE: Improving Textual Inversion for Zero-Shot Composed Image Retrieval
MagicLens (CLIP B)	23.8	MagicLens: Self-Supervised Image Retrieval with Open-Ended Instructions
CompoDiff (CLIP G/14)	17.71	CompoDiff: Versatile Composed Image Retrieval With Latent Diffusion
Context-I2W	14.62	Context-I2W: Mapping Images to Context-dependent Words for Accurate Zero-Shot Composed Image Retrieval
LDRE (CLIP L/14)	24.03	LDRE: LLM-based Divergent Reasoning and Ensemble for Zero-Shot Composed Image Retrieval
LinCIR (CLIP G/14)	21.01	Language-only Efficient Training of Zero-shot Composed Image Retrieval
CoLLM (Pretrained - BLIP-L/16)	20.4	-	-
CIReVL (CLIP B/32)	15.42	Vision-by-Language for Training-Free Compositional Image Retrieval
Pic2Word	9.51	Pic2Word: Mapping Pictures to Words for Zero-shot Composed Image Retrieval
CoVR-BLIP-2	29.55	CoVR-2: Automatic Data Construction for Composed Video Retrieval
OSrCIR (CLIP L/14)	25.33	Reason-before-Retrieve: One-Stage Reflective Chain-of-Thoughts for Training-Free Zero-Shot Composed Image Retrieval
MTCIR (BLIP B/16)	8.03	Pretrain like Your Inference: Masked Tuning Improves Zero-Shot Composed Image Retrieval	-
SEIZE (CLIP L/14)	25.82	Semantic Editing Increment Benefits Zero-Shot Composed Image Retrieval
MTCIR (CLIP L/14)	11.63	Pretrain like Your Inference: Masked Tuning Improves Zero-Shot Composed Image Retrieval	-
MagicLens (CoCa B)	32.0	MagicLens: Self-Supervised Image Retrieval with Open-Ended Instructions
OSrCIR (CLIP G/14)	31.14	Reason-before-Retrieve: One-Stage Reflective Chain-of-Thoughts for Training-Free Zero-Shot Composed Image Retrieval
MMRet-Base (CLIP B/16)	35.0	MegaPairs: Massive Data Synthesis For Universal Multimodal Retrieval
PALAVRA	5.32	"This is my unicorn, Fluffy": Personalizing frozen vision-language representations

0 of 42 row(s) selected.