SOTAVerified

Image Retrieval

Image Retrieval is a fundamental and long-standing computer vision task that involves finding images similar to a given query from a large database. It is often considered a form of fine-grained, instance-level classification. The task is integral to image recognition alongside classification and cross-modal retrieval. By leveraging visual similarity and other criteria, image retrieval enables users to efficiently discover relevant images, making it a crucial tool in applications such as search and recommendation.

Extending CLIP for Category-to-image Retrieval in E-commerce

( Image credit: DELF )

Papers

Recently AddedMost HypedMost ActiveNeeds VerificationMost Verified

Showing 1–50 of 2239 papers

TitleStatusHype
Visual Instruction TuningCode6
DINOv2: Learning Robust Visual Features without SupervisionCode6
GPT-4 Technical ReportCode6
Chinese CLIP: Contrastive Vision-Language Pretraining in ChineseCode5
CogVLM: Visual Expert for Pretrained Language ModelsCode5
Evaluating Pre-trained Convolutional Neural Networks and Foundation Models as Feature Extractors for Content-based Medical Image RetrievalCode4
Long-CLIP: Unlocking the Long-Text Capability of CLIPCode4
BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image Encoders and Large Language ModelsCode4
mPLUG-2: A Modularized Multi-modal Foundation Model Across Text, Image and VideoCode4
AltCLIP: Altering the Language Encoder in CLIP for Extended Language CapabilitiesCode4
Efficient LoFTR: Semi-Dense Local Feature Matching with Sparse-Like SpeedCode4
MegaPairs: Massive Data Synthesis For Universal Multimodal RetrievalCode3
MagicLens: Self-Supervised Image Retrieval with Open-Ended InstructionsCode3
A Comprehensive Survey on Composed Image RetrievalCode3
Gemini 1.5: Unlocking multimodal understanding across millions of tokens of contextCode3
MindEye2: Shared-Subject Models Enable fMRI-To-Image With 1 Hour of DataCode2
Linguistic-Aware Patch Slimming Framework for Fine-grained Cross-Modal AlignmentCode2
MixVPR: Feature Mixing for Visual Place RecognitionCode2
LDRE: LLM-based Divergent Reasoning and Ensemble for Zero-Shot Composed Image RetrievalCode2
INQUIRE: A Natural World Text-to-Image Retrieval BenchmarkCode2
iSEARLE: Improving Textual Inversion for Zero-Shot Composed Image RetrievalCode2
Learning Embeddings with Centroid Triplet Loss for Object Identification in Robotic GraspingCode2
Global Features are All You Need for Image Retrieval and RerankingCode2
Investigating the Role of Image Retrieval for Visual Localization -- An exhaustive benchmarkCode2
Bridging the Vision-Brain Gap with an Uncertainty-Aware Blur PriorCode2
Language-only Training of Zero-shot Composed Image RetrievalCode2
Fine-grained Image Captioning with CLIP RewardCode2
Local Feature Matching Using Deep Learning: A SurveyCode2
FastReID: A Pytorch Toolbox for General Instance Re-identificationCode2
Exploring a Fine-Grained Multiscale Method for Cross-Modal Remote Sensing Image RetrievalCode2
EMR-Merging: Tuning-Free High-Performance Model MergingCode2
Exploring the best way for UAV visual localization under Low-altitude Multi-view Observation Condition: a BenchmarkCode2
Grounding Language Models to Images for Multimodal Inputs and OutputsCode2
D3still: Decoupled Differential Distillation for Asymmetric Image RetrievalCode2
All You Need to Know About Training Image Retrieval ModelsCode2
Cross-view image geo-localization with Panorama-BEV Co-Retrieval NetworkCode2
EarthLoc: Astronaut Photography Localization by Indexing Earth from SpaceCode2
Composed Image Retrieval for Remote SensingCode2
Beyond Text: Optimizing RAG with Multimodal Inputs for Industrial ApplicationsCode2
Efficient Remote Sensing with Harmonized Transfer Learning and Modality AlignmentCode2
Encrypted Vector Similarity Computations Using Partially Homomorphic Encryption: Applications and Performance AnalysisCode2
Enhancing Diagnostic Accuracy in Rare and Common Fundus Diseases with a Knowledge-Rich Vision-Language ModelCode2
Composed Multi-modal Retrieval: A Survey of Approaches and ApplicationsCode2
CLEAR: A Fully User-side Image Search SystemCode2
Generating Images with Multimodal Language ModelsCode2
GeoCLIP: Clip-Inspired Alignment between Locations and Images for Effective Worldwide Geo-localizationCode2
AnyLoc: Towards Universal Visual Place RecognitionCode2
Cross the Gap: Exposing the Intra-modal Misalignment in CLIP via Modality InversionCode2
InstructBLIP: Towards General-purpose Vision-Language Models with Instruction TuningCode2
Image Textualization: An Automatic Framework for Creating Accurate and Detailed Image DescriptionsCode2
Show:102550
← PrevPage 1 of 45Next →
All datasetsROxford (Hard)ROxford (Medium)RParis (Hard)RParis (Medium)CREPE (Compositional REPresentation Evaluation)Fashion IQFlickr30K 1K testCIRRSOPFlickr30k-CNOxf5kFlickr30k

Benchmark Results

#ModelMetricClaimedVerifiedStatus
1SuperGlobalmAP80.2—Unverified
2AMESmAP80—Unverified
3Hypergraph propagation+community selectionmAP73—Unverified
4TokenmAP66.57—Unverified
5DELG+ α QE reranking+ RRT rerankingmAP64—Unverified
6FIRemAP61.2—Unverified
7HOWmAP56.9—Unverified
8ResNet101+ArcFace GLDv2-train-cleanmAP51.6—Unverified
9DELF–HQE+SPmAP50.3—Unverified
10HesAff–rSIFT–HQE+SPmAP49.7—Unverified
#ModelMetricClaimedVerifiedStatus
1AMESmAP90.7—Unverified
2Hypergraph propagation+Community selectionmAP88.4—Unverified
3TokenmAP82.28—Unverified
4FIRemAP81.8—Unverified
5DELG+ α QE reranking + RRT rerankingmAP80.4—Unverified
6HOWmAP79.4—Unverified
7ResNet101+ArcFace GLDv2-train-cleanmAP74.2—Unverified
8DELF–HQE+SPmAP73.4—Unverified
9HesAff–rSIFT–HQE+SPmAP71.3—Unverified
10DELF–ASMK*+SPmAP67.8—Unverified
#ModelMetricClaimedVerifiedStatus
1AMESmAP89.7—Unverified
2SuperGlobalmAP86.7—Unverified
3Hypergraph propagationmAP83.3—Unverified
4TokenmAP78.56—Unverified
5DELG+ α QE reranking + RRT rerankingmAP77.7—Unverified
6ResNet101+ArcFace GLDv2-train-cleanmAP70.3—Unverified
7FIRemAP70—Unverified
8DELF–HQE+SPmAP69.3—Unverified
9HOWmAP62.4—Unverified
10R–R-MACmAP59.4—Unverified
#ModelMetricClaimedVerifiedStatus
1AMESmAP94.9—Unverified
2Hypergraph propagationmAP92.6—Unverified
3TokenmAP89.34—Unverified
4DELG+ α QE reranking + RRT rerankingmAP88.5—Unverified
5FIRemAP85.3—Unverified
6ResNet101+ArcFace GLDv2-train-cleanmAP84.9—Unverified
7DELF–HQE+SPmAP84—Unverified
8HOWmAP81.6—Unverified
9R–R-MACmAP78.9—Unverified
10R–GeMmAP77.2—Unverified
#ModelMetricClaimedVerifiedStatus
1Swin-T (MosaiCLIP, CC-12M)Recall@1 (HN-Atom, UC)44.5—Unverified
2RN-50 (MosaiCLIP, CC-12M)Recall@1 (HN-Atom, UC)44.4—Unverified
3MosaiCLIP (YFCC-FT)Recall@1 (HN-Atom, UC)41.5—Unverified
4RN-50 (NegCLIP, CC-12M)Recall@1 (HN-Atom, UC)41.4—Unverified
5MosaiCLIP (CC-FT)Recall@1 (HN-Atom, UC)40.9—Unverified
6Swin-T (NegCLIP, CC-12M)Recall@1 (HN-Atom, UC)39.6—Unverified
7CLIP (YFCC-FT)Recall@1 (HN-Atom, UC)39.5—Unverified
8ViT-L-14 (LAION400M)Recall@1 (HN-Atom + HN-Comp, SC)39.44—Unverified
9NegCLIP (YFCC-FT)Recall@1 (HN-Atom, UC)39—Unverified
10CLIP-FT (YFCC-FT)Recall@1 (HN-Atom, UC)38.3—Unverified
#ModelMetricClaimedVerifiedStatus
1DQU-CIR(Recall@10+Recall@50)/271.77—Unverified
2TMCIR(Recall@10+Recall@50)/266.56—Unverified
3SPN4CIR (SPRC)(Recall@10+Recall@50)/266.41—Unverified
4SPRC(Recall@10+Recall@50)/264.85—Unverified
5Candidate Set Re-ranking(Recall@10+Recall@50)/262.15—Unverified
6RUTIR (BLIP B/16)(Recall@10+Recall@50)/261.32—Unverified
7CASE(Recall@10+Recall@50)/259.73—Unverified
8CaLa(Recall@10+Recall@50)/257.96—Unverified
9BLIP4CIR+Bi(Recall@10+Recall@50)/255.4—Unverified
10CLIP4Cir (v3)(Recall@10+Recall@50)/255.36—Unverified
#ModelMetricClaimedVerifiedStatus
1X-VLM (base)R@186.9—Unverified
2RCARR@162.6—Unverified
3SGRAFR@158.5—Unverified
4VisualSpartaR@157.4—Unverified
5LGSGMR@157.4—Unverified
6TERAN MrSwR@156.5—Unverified
7TERAN Symm.R@155.7—Unverified
8VSRNR@154.7—Unverified
9CAMPR@151.5—Unverified
10SCAN i-tR@144—Unverified
#ModelMetricClaimedVerifiedStatus
1TMCIR(Recall@5+Recall_subset@1)/283.46—Unverified
2SPN4CIR (SPRC)(Recall@5+Recall_subset@1)/282.69—Unverified
3SPRC2(Recall@5+Recall_subset@1)/282.66—Unverified
4SPRC(Recall@5+Recall_subset@1)/281.39—Unverified
5Candidate Set Re-ranking(Recall@5+Recall_subset@1)/280.9—Unverified
6CaLa(Recall@5+Recall_subset@1)/278.74—Unverified
7CASE (Pre-trained on LaSCo.Ca)(Recall@5+Recall_subset@1)/278.25—Unverified
8CASE(Recall@5+Recall_subset@1)/277.5—Unverified
9VISTA (base)(Recall@5+Recall_subset@1)/275.9—Unverified
10MMRet-MLLM(Recall@5+Recall_subset@1)/275.7—Unverified
#ModelMetricClaimedVerifiedStatus
1Unicom+ViT-L@336pxR@191.2—Unverified
2ROADMAP (DeiT-B)R@186—Unverified
3CGD (SG/GS)R@184.2—Unverified
4ROADMAP (ResNet-50)R@183.1—Unverified
5ProxyNCA++R@181.4—Unverified
6PNP LossR@181.1—Unverified
7Cross-Batch MemoryR@180.6—Unverified
8Smooth-APR@180.1—Unverified
9NormSoftmax2048 (ResNet-50)R@179.5—Unverified
10EPSHN512R@178.3—Unverified
#ModelMetricClaimedVerifiedStatus
1InternVL-G-FTR@185.9—Unverified
2InternVL-C-FTR@185.2—Unverified
3R2D2 (ViT-L/14)R@184.4—Unverified
4CN-CLIP (ViT-L/14@336px)R@184.4—Unverified
5CN-CLIP (ViT-H/14)R@183.8—Unverified
6CN-CLIP (ViT-L/14)R@182.7—Unverified
7CN-CLIP (ViT-B/16)R@179.1—Unverified
8R2D2 (ViT-B)R@178.3—Unverified
9Wukong (ViT-L/14)R@177.4—Unverified
10Wukong (ViT-B/32)R@167.6—Unverified
#ModelMetricClaimedVerifiedStatus
1Offline DiffusionMAP96.2—Unverified
2CNN+IME layerMAP92—Unverified
3DELF+FT+ATT+DIR+QEMAP90—Unverified
4DIR+QE*MAP89—Unverified