# Deeply learned face representations are sparse, selective, and robust

arXiv: https://arxiv.org/abs/1412.1265 · HTML: https://mlsanity.com/paper/1412.1265/

## Key facts

- Cited by 56 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 3 Dec 2014.

## Built on by (most cited first, 30 of 56)

- [Feature Importance-aware Transferable Adversarial Attacks](https://mlsanity.com/paper/2107.14185/) (2021)
- [Few-Shot Audio-Visual Learning of Environment Acoustics](https://mlsanity.com/paper/2206.04006/) (2022)
- [SphereFace2: Binary Classification is All You Need for Deep Face Recognition](https://mlsanity.com/paper/2108.01513/) (2021)
- [WebFace260M: A Benchmark for Million-Scale Deep Face Recognition](https://mlsanity.com/paper/2204.10149/) (2022)
- [Move2Hear: Active Audio-Visual Source Separation](https://mlsanity.com/paper/2105.07142/) (2021)
- [Federated Learning for Face Recognition with Gradient Correction](https://mlsanity.com/paper/2112.07246/) (2021)
- [DeepFace-EMD: Re-ranking Using Patch-wise Earth Mover's Distance Improves Out-Of-Distribution Face Identification](https://mlsanity.com/paper/2112.04016/) (2021)
- [Towards Accurate and Compact Architectures via Neural Architecture Transformer](https://mlsanity.com/paper/2102.10301/) (2021)
- [Active Audio-Visual Separation of Dynamic Sound Sources](https://mlsanity.com/paper/2202.00850/) (2022)
- [Retrieve in Style: Unsupervised Facial Feature Transfer and Retrieval](https://mlsanity.com/paper/2107.06256/) (2021)
- [50 Years of Automated Face Recognition](https://mlsanity.com/paper/2505.24247/) (2025)
- [Searching for Alignment in Face Recognition](https://mlsanity.com/paper/2102.05447/) (2021)
- [An Efficient Training Approach for Very Large Scale Face Recognition](https://mlsanity.com/paper/2105.10375/) (2021)
- [Deep Rival Penalized Competitive Learning for Low-resolution Face Recognition](https://mlsanity.com/paper/2108.01286/) (2021)
- [Deep Learning for Scene Classification: A Survey](https://mlsanity.com/paper/2101.10531/) (2021)
- [Chat2Map: Efficient Scene Mapping from Multi-Ego Conversations](https://mlsanity.com/paper/2301.02184/) (2023)
- [MorDeephy: Face Morphing Detection Via Fused Classification](https://mlsanity.com/paper/2208.03110/) (2022)
- [CoReFace: Sample-Guided Contrastive Regularization for Deep Face Recognition](https://mlsanity.com/paper/2304.11668/) (2023)
- [Improved weight initialization for deep and narrow feedforward neural network](https://mlsanity.com/paper/2311.03733/) (2023)
- [Deep Adaptive Arbitrary Polynomial Chaos Expansion: A Mini-data-driven Semi-supervised Method for Uncertainty Quantification](https://mlsanity.com/paper/2107.10428/) (2021)
- [Recent Advances in Deep Learning Techniques for Face Recognition](https://mlsanity.com/paper/2103.10492/) (2021)
- [Saliency-Guided Hidden Associative Replay for Continual Learning](https://mlsanity.com/paper/2310.04334/) (2023)
- [Content-Aware Convolutional Neural Networks](https://mlsanity.com/paper/2106.15797/) (2021)
- [Mask or Non-Mask? Robust Face Mask Detector via Triplet-Consistency Representation Learning](https://mlsanity.com/paper/2110.00523/) (2021)
- [PAM: Pose Attention Module for Pose-Invariant Face Recognition](https://mlsanity.com/paper/2111.11940/) (2021)
- [Unified Negative Pair Generation toward Well-discriminative Feature Space for Face Recognition](https://mlsanity.com/paper/2203.11593/) (2022)
- [Deep face recognition with clustering based domain adaptation](https://mlsanity.com/paper/2205.13937/) (2022)
- [Towards Robust Face Recognition with Comprehensive Search](https://mlsanity.com/paper/2208.13600/) (2022)
- [Young Labeled Faces in the Wild (YLFW): A Dataset for Children Faces Recognition](https://mlsanity.com/paper/2301.05776/) (2023)
- [Inter-class Discrepancy Alignment for Face Recognition](https://mlsanity.com/paper/2103.01559/) (2021)

## Abstract (arXiv)

This paper designs a high-performance deep convolutional network (DeepID2+) for face recognition. It is learned with the identification-verification supervisory signal. By increasing the dimension of hidden representations and adding supervision to early convolutional layers, DeepID2+ achieves new state-of-the-art on LFW and YouTube Faces benchmarks. Through empirical studies, we have discovered three properties of its deep neural activations critical for the high performance: sparsity, selectiveness and robustness. (1) It is observed that neural activations are moderately sparse. Moderate sparsity maximizes the discriminative power of the deep net as well as the distance between images. It is surprising that DeepID2+ still can achieve high recognition accuracy even after the neural responses are binarized. (2) Its neurons in higher layers are highly selective to identities and identity-related attributes. We can identify different subsets of neurons which are either constantly excited or inhibited when different identities or attributes are present. Although DeepID2+ is not taught to distinguish attributes during training, it has implicitly learned such high-level concepts. (3) It is much more robust to occlusions, although occlusion patterns are not included in the training set.