mlsanityData to 14 Oct 2026

Asymmetric Image Retrieval with Cross Model Compatible Ensembles

arXiv 2303.17531 ↗First posted 30 Mar 2023cs.CVcited by 2

On the map: Knowledge distillation → General knowledge distillation methods

Abstract (arXiv)

The asymmetrical retrieval setting is a well suited solution for resource constrained applications such as face recognition and image retrieval. In this setting, a large model is used for indexing the gallery while a lightweight model is used for querying. The key principle in such systems is ensuring that both models share the same embedding space. Most methods in this domain are based on knowledge distillation. While useful, they suffer from several drawbacks: they are upper-bounded by the performance of the single best model found and cannot be extended to use an ensemble of models in a straightforward manner. In this paper we present an approach that does not rely on knowledge distillation, rather it utilizes embedding transformation models. This allows the use of N independently trained and diverse gallery models (e.g., trained on different datasets or having a different architecture) and a single query model. As a result, we improve the overall accuracy beyond that of any single model while maintaining a low computational budget for querying. Additionally, we propose a gallery image rejection method that utilizes the diversity between multiple transformed embeddings to estimate the uncertainty of gallery images.

Key facts. Cited by 2 papers in the map (data through 14 Oct 2026). Placed in General knowledge distillation methods, one of 12 directions in Knowledge distillation. Builds on 30 of its 61 references that are in the map. First posted on arXiv 30 Mar 2023.

Builds on

30 of 61 references.

References · 30

  1. 2015Deep Residual Learning for Image Recognitioncited by 52,096
  2. 2014Very Deep Convolutional Networks for Large-Scale Image Recognitioncited by 15,364
  3. 2015Distilling the Knowledge in a Neural Networkcited by 10,324
  4. 2014Going Deeper with Convolutionscited by 4,241
  5. 2016Simple and Scalable Predictive Uncertainty Estimation using Deep Ensemblescited by 3,926
  6. 2015FaceNet: A Unified Embedding for Face Recognition and Clusteringcited by 2,849
  7. 2019Similarity of Neural Network Representations Revisitedcited by 1,755
  8. 2018CosFace: Large Margin Cosine Loss for Deep Face Recognitioncited by 971
  9. 2017VGGFace2: A dataset for recognising faces across pose and agecited by 914
  10. 2017Selective Classification for Deep Neural Networkscited by 852
  11. 2017Fine-tuning CNN Image Retrieval with No Human Annotationcited by 499
  12. 2015Deep Metric Learning via Lifted Structured Feature Embeddingcited by 449
18 more
  1. 2014Deep metric learning using Triplet networkcited by 422
  2. 2020Circle Loss: A Unified Perspective of Pair Similarity Optimizationcited by 403
  3. 2018Additive Margin Softmax for Face Verificationcited by 278
  4. 2017NormFace: L2 Hypersphere Embedding for Face Verificationcited by 237
  5. 2014Deep Learning Face Representation by Joint Identification-Verificationcited by 162
  6. 2020Towards Backward-Compatible Representation Learningcited by 75
  7. 2018Towards Understanding Learning Representations: To What Extent Do Different Neural Networks Learn the Same Representationcited by 62
  8. 2020Learning and aggregating deep local descriptors for instance-level recognitioncited by 55
  9. 2017The Relative Performance of Ensemble Methods with Deep Convolutional Neural Networks for Image Classificationcited by 50
  10. 2014Visual Instance Retrieval with Deep Convolutional Networkscited by 48
  11. 2019Local Features and Visual Words Emerge in Activationscited by 37
  12. 2021Learning Compatible Embeddingscited by 37
  13. 2020Unified Representation Learning for Cross Model Compatibilitycited by 30
  14. 2020Asymmetric metric learning for knowledge transfercited by 29
  15. 2021Compatibility-aware Heterogeneous Visual Searchcited by 23
  16. 2022Towards Universal Backward-Compatible Representation Learningcited by 21
  17. 2022BT2: Backward-compatible Training with Basis Transformationcited by 8
  18. 2023FastFill: Efficient Compatible Model Updatecited by 7

Built on by

Cited by year

Citing papers per year: 2024: 1; 2025: 1.

2 papers.

  1. 2024Data Pruning via Separability, Integrity, and Model Uncertainty-Aware Importance Samplingcited by 2
  2. 2025Learning Compatible Multi-Prize Subnetworks for Asymmetric Retrievalcited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2024Structure Similarity Preservation Learning for Asymmetric Image Retrieval
  2. 2025QuARI: Query Adaptive Retrieval Improvement
  3. 2026MASCOT: Model-Aware Submodular Coverage for Composite-Attribute Text-to-Image Retrieval
  4. 2024Image2Sentence based Asymmetrical Zero-shot Composed Image Retrieval
  5. 2023SynCDR : Training Cross Domain Retrieval Models with Synthetic Data
  6. 2023Exploiting Unlabelled Photos for Stronger Fine-Grained SBIR
  7. 2023EmbedDistill: A Geometric Knowledge Distillation for Information Retrieval
  8. 2024NUDGE: Lightweight Non-Parametric Fine-Tuning of Embeddings for Retrieval