# A Benchmark and Asymmetrical-Similarity Learning for Practical Image Copy Detection

arXiv: https://arxiv.org/abs/2205.12358 · HTML: https://mlsanity.com/paper/2205.12358/

## Key facts

- Cited by 9 papers in the map (data through 14 Oct 2026).
- Builds on 16 of its 29 references that are in the map.
- First posted on arXiv 24 May 2022.

## Builds on (most cited first, 16 of 16)

- [A Simple Framework for Contrastive Learning of Visual Representations](https://mlsanity.com/paper/2002.05709/) (2020)
- [Momentum Contrast for Unsupervised Visual Representation Learning](https://mlsanity.com/paper/1911.05722/) (2019)
- [Bootstrap your own latent: A new approach to self-supervised Learning](https://mlsanity.com/paper/2006.07733/) (2020)
- [CosFace: Large Margin Cosine Loss for Deep Face Recognition](https://mlsanity.com/paper/1801.09414/) (2018)
- [In Defense of the Triplet Loss for Person Re-Identification](https://mlsanity.com/paper/1703.07737/) (2017)
- [Circle Loss: A Unified Perspective of Pair Similarity Optimization](https://mlsanity.com/paper/2002.10857/) (2020)
- [MagFace: A Universal Representation for Face Recognition and Quality Assessment](https://mlsanity.com/paper/2103.06627/) (2021)
- [Large-Margin Softmax Loss for Convolutional Neural Networks](https://mlsanity.com/paper/1612.02295/) (2016)
- [A Self-Supervised Descriptor for Image Copy Detection](https://mlsanity.com/paper/2202.10261/) (2022)
- [MultiGrain: a unified image embedding for classes and instances](https://mlsanity.com/paper/1902.05509/) (2019)
- [The 2021 Image Similarity Dataset and Challenge](https://mlsanity.com/paper/2106.09672/) (2021)
- [Contrastive Learning with Large Memory Bank and Negative Embedding Subtraction for Accurate Copy Detection](https://mlsanity.com/paper/2112.04323/) (2021)
- [Bag of Tricks and A Strong baseline for Image Copy Detection](https://mlsanity.com/paper/2111.08004/) (2021)
- [D2LV: A Data-Driven and Local-Verification Approach for Image Copy Detection](https://mlsanity.com/paper/2111.07090/) (2021)
- [Producing augmentation-invariant embeddings from real-life imagery](https://mlsanity.com/paper/2112.03415/) (2021)
- [3rd Place: A Global and Local Dual Retrieval Solution to Facebook AI Image Similarity Challenge](https://mlsanity.com/paper/2112.02373/) (2021)

## Built on by (most cited first, 9 of 9)

- [TIP-I2V: A Million-Scale Real Text and Image Prompt Dataset for Image-to-Video Generation](https://mlsanity.com/paper/2411.04709/) (2024)
- [Active Image Indexing](https://mlsanity.com/paper/2210.10620/) (2022)
- [AnyPattern: Towards In-context Image Copy Detection](https://mlsanity.com/paper/2404.13788/) (2024)
- [Image Copy Detection for Diffusion Models](https://mlsanity.com/paper/2409.19952/) (2024)
- [Feature-compatible Progressive Learning for Video Copy Detection](https://mlsanity.com/paper/2304.10305/) (2023)
- [3rd Place Solution to Meta AI Video Similarity Challenge](https://mlsanity.com/paper/2304.11964/) (2023)
- [Relational Self-supervised Distillation with Compact Descriptors for Image Copy Detection](https://mlsanity.com/paper/2405.17928/) (2024)
- [Origin Identification for Text-Guided Image-to-Image Diffusion Models](https://mlsanity.com/paper/2501.02376/) (2025)
- [Tracing Copied Pixels and Regularizing Patch Affinity in Copy Detection](https://mlsanity.com/paper/2602.17484/) (2026)

## Abstract (arXiv)

Image copy detection (ICD) aims to determine whether a query image is an edited copy of any image from a reference set. Currently, there are very limited public benchmarks for ICD, while all overlook a critical challenge in real-world applications, i.e., the distraction from hard negative queries. Specifically, some queries are not edited copies but are inherently similar to some reference images. These hard negative queries are easily false recognized as edited copies, significantly compromising the ICD accuracy. This observation motivates us to build the first ICD benchmark featuring this characteristic. Based on existing ICD datasets, this paper constructs a new dataset by additionally adding 100, 000 and 24, 252 hard negative pairs into the training and test set, respectively. Moreover, this paper further reveals a unique difficulty for solving the hard negative problem in ICD, i.e., there is a fundamental conflict between current metric learning and ICD. This conflict is: the metric learning adopts symmetric distance while the edited copy is an asymmetric (unidirectional) process, e.g., a partial crop is close to its holistic reference image and is an edited copy, while the latter cannot be the edited copy of the former (in spite the distance is equally small). This insight results in an Asymmetrical-Similarity Learning (ASL) method, which allows the similarity in two directions (the query <-> the reference image) to be different from each other. Experimental results show that ASL outperforms state-of-the-art methods by a clear margin, confirming that solving the symmetric-asymmetric conflict is critical for ICD. The NDEC dataset and code are available at https://github.com/WangWenhao0716/ASL.