# Rethinking Feature Discrimination and Polymerization for Large-scale Recognition

arXiv: https://arxiv.org/abs/1710.00870 · HTML: https://mlsanity.com/paper/1710.00870/

## Key facts

- Cited by 13 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 2 Oct 2017.

## Built on by (most cited first, 13 of 13)

- [SphereFace Revived: Unifying Hyperspherical Face Recognition](https://mlsanity.com/paper/2109.05565/) (2021)
- [Learning to Disentangle Scenes for Person Re-identification](https://mlsanity.com/paper/2111.05476/) (2021)
- [IronMask: Modular Architecture for Protecting Deep Face Template](https://mlsanity.com/paper/2104.02239/) (2021)
- [Generalized Inter-class Loss for Gait Recognition](https://mlsanity.com/paper/2210.06779/) (2022)
- [Fine-Grained VR Sketching: Dataset and Insights](https://mlsanity.com/paper/2209.10008/) (2022)
- [Fine-Grained Visual Classification via Simultaneously Learning of Multi-regional Multi-grained Features](https://mlsanity.com/paper/2102.00367/) (2021)
- [Prototype Memory for Large-scale Face Representation Learning](https://mlsanity.com/paper/2105.02103/) (2021)
- [A Deep Positive-Negative Prototype Approach to Integrated Prototypical Discriminative Learning](https://mlsanity.com/paper/2501.02477/) (2025)
- [Sphere Face Model:A 3D Morphable Model with Hypersphere Manifold Latent Space](https://mlsanity.com/paper/2112.02238/) (2021)
- [Profile to Frontal Face Recognition in the Wild Using Coupled Conditional GAN](https://mlsanity.com/paper/2107.13742/) (2021)
- [Rapid detection and recognition of whole brain activity in a freely behaving Caenorhabditis elegans](https://mlsanity.com/paper/2109.10474/) (2021)
- [Fine-tune your Classifier: Finding Correlations With Temperature](https://mlsanity.com/paper/2210.09715/) (2022)
- [Improving Generalization in Deepfake Detection with Face Foundation Models and Metric Learning](https://mlsanity.com/paper/2508.19730/) (2025)

## Abstract (arXiv)

Feature matters. How to train a deep network to acquire discriminative features across categories and polymerized features within classes has always been at the core of many computer vision tasks, specially for large-scale recognition systems where test identities are unseen during training and the number of classes could be at million scale. In this paper, we address this problem based on the simple intuition that the cosine distance of features in high-dimensional space should be close enough within one class and far away across categories. To this end, we proposed the congenerous cosine (COCO) algorithm to simultaneously optimize the cosine similarity among data. It inherits the softmax property to make inter-class features discriminative as well as shares the idea of class centroid in metric learning. Unlike previous work where the center is a temporal, statistical variable within one mini-batch during training, the formulated centroid is responsible for clustering inner-class features to enforce them polymerized around the network truncus. COCO is bundled with discriminative training and learned end-to-end with stable convergence. Experiments on five benchmarks have been extensively conducted to verify the effectiveness of our approach on both small-scale classification task and large-scale human recognition problem.