Clothes image retrieval is a task that retrieves exact or very similar clothes from a large clothes image gallery according to a given query image. The high deformability of clothes and similar designs make clothes retrieval very difficult. Convolutional neural networks perform well in the feature extraction field, but cannot extract sufficiently fine-grained features for clothes retrieval tasks. Therefore, a model based on ResNet was designed in this article, with several improvements made to it to ensure better feature discrimination. On the other hand, pair-based loss functions that are commonly used in image retrieval suffer from insufficient discrimination ability, difficulty in convergence, and time-consuming during training. Cluster triplet loss function is proposed in this article, which reduces the computational complexity in the training phase and enables the model to have a certain ability to resist noise labels. Experiments are completed on In-shop clothes retrieval benchmark and Consumer-to-shop clothes retrieval benchmark in Deepfashion, and stanford online products (SOP) to supplement. Our method can have Recall@1 improvements of 1.3–1.7% on all three datasets, surpassing the latest state-of-the-art methods.
SzegedyCLiuWJiaY, et al.
Going deeper with convolutions. Proceedings of the IEEE conference on computer vision and pattern recognition. 07–12 June, 2015. IEEE: Boston, MA, USA. pp. 1–9. Doi: 10.1109/CVPR.2015.7298594
2.
SimonyanKZissermanA.Very deep convolutional networks for large-scale image recognition. International conference on learning representations. 14–16 April, 2014, arXiv: Banff, Canada. 2015. 1–14.
3.
HeKZhangXRenS, et al.
Deep residual learning for image recognition. Proceedings of the IEEE conference on computer vision and pattern recognition. 27–30 June, 2016. IEEE: Las Vegas, NV, USA, pp. 770–778. Doi: 10.1109/CVPR.2016.90
4.
ZhouZLiuMDengW, et al.
Clothing image classification algorithm based on convolutional neural network and optimized regularized extreme learning machine. Text Res J2022; 92: 23–24.
5.
ZhouZDengWWangY, et al.
Classification of clothing images based on a parallel convolutional neural network and random vector functional link optimized by the grasshopper optimization algorithm. Text Res J2022;
92: 1415–1428.
6.
LiJShiWYangD.Clothing image classification with a dragonfly algorithm optimised online sequential extreme learning machine. Fibres Text East Eur2021; 29: 91–96.
HadsellRChopraSLeCunY.Dimensionality reduction by learning an invariant mapping. Proceedings of the IEEE conference on computer vision and pattern recognition. 17–22 June, 2006. IEEE: Las New York, NY, USA. pp. 1735–1742. Doi: 10.1109/CVPR.2006.100
9.
SchroffFKalenichenkoDPhilbinJ.Facenet: A unified embedding for face recognition and clustering. Proceedings of the IEEE conference on computer vision and pattern recognition. 07–12 June, 2015. IEEE: Boston, MA, USA. pp. 815–823. Doi: 10.1109/CVPR.2015.7298682
10.
DalalNTriggsB.Histograms of oriented gradients for human detection. Proceedings of the IEEE conference on computer vision and pattern recognition. 20–25 June, 2005. IEEE: San Diego, CA, USA. pp. 886–893. Doi: 10.1109/CVPR.2005.177
11.
TolaELepetitVFuaP.DAISY: An efficient dense descriptor applied to wide-baseline stereo. IEEE Trans Pami2010;
32: 815–830.
12.
LoweDG.Distinctive image features from scale-invariant keypoints. Int J Comput Vision2004;
60: 91–110.
13.
RazavianASAzizpourHSullivanJ, et al.
CNN features off-the-shelf: An astounding baseline for recognition. Proceedings of the international conference on computer vision and pattern recognition workshops 23–28 June, 2014. IEEE: Columbus, OH, USA. pp. 806–813. Doi: 10.1109/CVPRW.2014.131
14.
GordoAAlmazánJRevaudJ, et al.
End-to-end learning of deep visual representations for image retrieval. Int J Comput Vision2017;
124: 237–254.
15.
WangXZhangHHuangW, et al.
Cross-batch memory for embedding learning.Proceedings of the international conference on computer vision and pattern recognition 13-19 June, 2020. IEEE: Seattle, WA, USA. pp. 6388–6397. Doi: 10.1109/CVPR42600.2020.00642
16.
Movshovitz-AttiasYToshevALeungTK, et al.
No fuss distance metric learning using proxies. Proceedings of the IEEE/CVF international conference on computer vision. 22–29 October, 2017. IEEE: Seattle, Venice, Italy. 360–368. Doi: 10.1109/ICCV.2017.47
17.
VenkataramananSPsomasBKijakE, et al.
It takes two to tango: Mixup for deep metric learning. International conference on learning representations. 25-29 April, 2022. Springer, Cham: Virtual.
18.
KimSKimDChoM, et al.
Proxy anchor loss for deep metric learning. Proceedings of the IEEE conference on computer vision and pattern recognition 13–19 June, 2020, IEEE: Seattle, WA, USA. pp. 3235–3244. Doi: 10.1109/CVPR42600.2020.00330
19.
WieczorekMRychalskaBDabrowskiJ.On the unreasonable effectiveness of centroids in image retrieval. Neural information processing 28th international conference 28 Novomeber-09 December, 2021 Publish on 05 December, 2021, Springer, Cham: New Orleans, LA, USA. pp. 212–223.
20.
LiuZLuoPQiuS, et al.
Deepfashion: Powering robust clothes recognition and retrieval with rich annotations. Proceedings of the IEEE conference on computer vision and pattern recognition 27–30 June, 2016. IEEE: Las Vegas, NV, USA. pp. 1096–1104. Doi: 10.1109/CVPR.2016.124
21.
SongHOXiangYJegelkaS, et al.
Deep metric learning via lifted structured feature embedding. Proceedings of the IEEE conference on computer vision and pattern recognition 27–30 June, 2016. IEEE: Las Vegas, NV, USA. pp. 4004–4012. Doi: 10.1109/CVPR.2016.434
22.
TehEWDeVriesTTaylorGW.Proxynca++: Revisiting and revitalizing proxy neighborhood component analysis. Proceedings of the European conference on computer vision 23–28 AUGUST, 2020. Glasgow, UK. pp. 448–464.
23.
KuangZGaoYLiG, et al.
Fashion retrieval via graph reasoning networks on a similarity pyramid. Proceedings of the IEEE/CVF international conference on computer vision 27 October – 02 November, 2019, IEEE: Seoul, Korea (South), 2019. pp. 3066–3075. Doi: 10.1109/ICCV.2019.00316
24.
WieczorekMMichalowskiAWróblewskaA, et al.
A strong baseline for fashion retrieval with person re-identification models. Neural information processing – 27th international conference 06–12 December, 2020, Springer, Cham: Virtual. pp. 294–301.
