Papers › Circle Loss: A Unified Perspective of Pair Similarity Optimization

Circle Loss: A Unified Perspective of Pair Similarity Optimization

25 Feb 2020CVPR 2020 6arXiv:2002.10857archive 2025-07-28

Yifan Sun, Changmao Cheng, Yuhan Zhang, Chi Zhang, Liang Zheng, Zhongdao Wang, Yichen Wei

This paper provides a pair similarity optimization viewpoint on deep feature learning, aiming to maximize the within-class similarity sₚ and minimize the between-class similarity sₙ. We find a majority of loss functions, including the triplet loss and the softmax plus cross-entropy loss, embed sₙ and sₚ into similarity pairs and seek to reduce (sₙ-sₚ). Such an optimization manner is inflexible, because the penalty strength on every single similarity score is restricted to be equal. Our intuition is that if a similarity score deviates far from the optimum, it should be emphasized. To this end, we simply re-weight each similarity to highlight the less-optimized similarity scores. It results in a Circle loss, which is named due to its circular decision boundary. The Circle loss has a unified formula for two elemental deep feature learning approaches, i.e. learning with class-level labels and pair-wise labels. Analytically, we show that the Circle loss offers a more flexible optimization approach towards a more definite convergence target, compared with the loss functions optimizing (sₙ-sₚ). Experimentally, we demonstrate the superiority of the Circle loss on a variety of deep feature learning tasks. On face recognition, person re-identification, as well as several fine-grained image retrieval datasets, the achieved performance is on par with the state of the art.

PaperPDFConference PDFCodeCode Syntology ran

In Syntology Open this paper in Syntology's Atlas, the map of the papers in Syntology's graph and their citations.

For agents, Syntology's MCP tool lists every function and class Syntology harvested from this paper and whether it ran (how to connect): get_harvested_code_for_paper(arxiv_id="2002.10857")

Code

Syntology Ran 2 of 2 code samples harvested from 2 repositories linked to this paper; 0 have no recorded run. Of those that ran: 1 ran · our draft was wrong; 1 ran · fixture could not drive it.

By repository: community (archive-listed): 2 samples from 2 repositories, 2 ran. The run record, sample by sample. “Ran” means executed on a synthesized input, not that the code is correct or reproduces the paper.

16 repositories listed; official and paper-mentioned ones first.

FEIfei-coder/circle-loss-for-reid mentioned on GitHubpytorch report
TinyZeaMays/CircleLoss mentioned on GitHubpytorch report
XuyangBai/D3Feat mentioned on GitHubtfMIT report
layumi/Person_reID_baseline_pytorch mentioned on GitHubpytorch report
lzx551402/ASLFeat mentioned on GitHubtfMIT report
qianjinhao/circle-loss mentioned on GitHubpytorchMIT report
wujpbb7/caffe_circleloss mentioned on GitHub report
yujiacheng333/CircleLossMNIST mentioned on GitHubtf report
zhen8838/Circle-Loss mentioned on GitHubtfMIT report

Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.

Code Syntology ran Syntology

2 samples harvested; 2 ran; 0 honoured the contract we drafted; 0 have no recorded run. Read from Syntology's graph 2026-09-24; that is when this build read the record, not when the samples ran.

1ran · our draft was wrong
1ran · fixture could not drive it

Licence: 1 of the 2 samples is pointer only, meaning Syntology does not serve that copy's text. This page shows no code text for any sample; each one links to its file in the repository.

Harvested from 2 repositories linked to this paper, official or community; each sample names its own and says which. “Ran” means the sample executed on a synthesized input. It does not mean the output is correct, and nothing here reproduces the paper's results. “Honoured” and “violated” refer to a contract Syntology drafted from the code itself; “our draft was wrong” and “fixture could not drive it” are failures of Syntology's instrument, not of the code.

Each sample ends with its code_sha256, Syntology's identity for that exact code. An agent fetches the stored sample with Syntology's MCP tool get_code(code_sha256="…") (how to connect); click an identity to copy that call.

Repository labels, per sample. official repository: The archive marks this repository official for the paper. named in the paper: The archive records that the paper mentions this repository; it is not marked official. community (archive-listed): In the archive's code links for this paper, not marked official and not recorded as mentioned in the paper. found in paper text by Syntology: Syntology found this repository in the paper's own text; whether it is the authors' implementation is not asserted. community: Not in the archive's code links for this paper; a community repository Syntology harvested. Samples from a repository marked official are listed first. Licence labels name the repository's licence as recorded at harvest. “Pointer only” means Syntology does not serve that copy's text, for one of four reasons: no licence file was found; the licence was not identified; the licence is recorded as permissive but that copy's record is not marked cleared; or the licence is outside the permissive list Syntology serves text under (MIT, Apache-2.0, BSD and similar). Some licences outside that list permit redistribution, such as WTFPL, and GPL-3.0 under its conditions; they are simply not on the list. Hover a licence label for the reason. File links open the file on GitHub at the default branch, which may have changed since the harvest.

DeepSupervision FEIfei-coder/circle-loss-for-reid/losses.py community (archive-listed) ran · our draft was wrong MIT (permissive) · b33437b2fbf7912a · report
convert_label_to_similarity TinyZeaMays/CircleLoss/circle_loss.py community (archive-listed) ran · fixture could not drive it fingerprinted no licence file found · pointer only · 05c14581c9d9da1d · report

Tasks

Face RecognitionFace VerificationImage RetrievalMetric LearningPerson Re-IdentificationRetrieval

1 archive task tag without a task page not shown.

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Face Recognition CFP-FP CircleLoss(ours) Accuracy 0.9602 #5 of 8 Archive leaderboard report
Face Recognition LFW CircleLoss Accuracy 0.9973 #9 of 16 Archive leaderboard report
Face Verification IJB-C circle loss TAR @ FAR=1e-3 96.29% #11 of 26 Archive leaderboard report
Face Verification IJB-C circle loss TAR @ FAR=1e-4 93.95% #11 of 26 Archive leaderboard report
Face Verification IJB-C circle loss TAR @ FAR=1e-5 89.60% #11 of 26 Archive leaderboard report
Face Verification IJB-C circle loss model R100 #11 of 26 Archive leaderboard report
Face Verification IJB-C circle loss training dataset MS1M Cleaned #11 of 26 Archive leaderboard report
Metric Learning CARS196 CircleLoss R@1 83.4 #30 of 36 Archive leaderboard report
Metric Learning Stanford Online Products Circle Loss R@1 78.3 #28 of 33 Archive leaderboard report
Person Re-Identification MSMT17 MGN + CircleLoss(ours) Rank-1 76.9 #35 of 43 Archive leaderboard report
Person Re-Identification MSMT17 MGN + CircleLoss(ours) mAP 52.1 #35 of 43 Archive leaderboard report
Person Re-Identification MSMT17 ResNet50 + CircleLoss(ours) Rank-1 76.3 #36 of 43 Archive leaderboard report
Person Re-Identification MSMT17 ResNet50 + CircleLoss(ours) mAP 50.2 #36 of 43 Archive leaderboard report
Person Re-Identification Market-1501 MGN + CircleLoss(ours) Rank-1 96.1 #32 of 135 Archive leaderboard report
Person Re-Identification Market-1501 MGN + CircleLoss(ours) mAP 87.4 #32 of 135 Archive leaderboard report
Person Re-Identification Market-1501 ResNet50 + CircleLoss(ours) Rank-1 94.2 #80 of 135 Archive leaderboard report
Person Re-Identification Market-1501 ResNet50 + CircleLoss(ours) mAP 84.9 #80 of 135 Archive leaderboard report

Ranks are positions in the archive's leaderboards as they stood at the 2025-07-28 snapshot. Results published since then are not among these rows, so a rank here is not a current standing.

Methods

SoftmaxTriplet Loss

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections