Recent face recognition experiments on the LFW benchmark show that face
recognition is performing stunningly well, surpassing human recognition rates.
In this paper, we study face recognition at scale. Specifically, we have
collected from Flickr a \textbf{Million} faces and evaluated state of the art
face recognition algorithms on this dataset. We found that the performance of
algorithms varies--while all perform great on LFW, once evaluated at scale
recognition rates drop drastically for most algorithms. Interestingly, deep
learning based approach by \cite{schroff2015facenet} performs much better, but
still gets less robust at scale. We consider both verification and
identification problems, and evaluate how pose affects recognition at scale.
Moreover, we ran an extensive human study on Mechanical Turk to evaluate human
recognition at scale, and report results. All the photos are creative commons
photos and is released at \small{http://megaface.cs.washington.edu/} for
research and further experiments.

研究者使用 Million 的人脸图像数据集，评估各种最先进的人脸识别算法的表现，并发现大规模情况下的性能变化，尤其是深度学习方法在大规模情况下依然表现良好，但对于规模的变化也变得不够鲁棒。研究者还通过机械土耳其进行了广泛人类研究，并报告了结果。

MegaFace: 百万级别人脸识别

MegaFace: A Million Faces for Recognition at Scale

We present a representation learning method that learns features at multiple
different levels of scale. Working within the unsupervised framework of
denoising autoencoders, we observe that when the input is heavily corrupted
during training, the network tends to learn coarse-grained features, whereas
when the input is only slightly corrupted, the network tends to learn
fine-grained features. This motivates the scheduled denoising autoencoder,
which starts with a high level of noise that lowers as training progresses. We
find that the resulting representation yields a significant boost on a later
supervised task compared to the original input, or to a standard denoising
autoencoder trained at a single noise level. After supervised fine-tuning our
best model achieves the lowest ever reported error on the CIFAR-10 data set
among permutation-invariant methods.

本文提出了一种表示学习方法，可以在多个不同的尺度下学习特征，在无监督的去噪自编码器框架内工作，观察到在训练期间当输入受到严重破坏时，网络倾向于学习粗粒度特征，而当输入仅稍微破坏时，网络倾向于学习细粒度特征。本文提出了一个定制的去噪自编码器，在训练过程中开始时使用高水平的噪声，随着训练的进行，降低噪声水平。我们发现所得到的表示可以在后续的监督任务上显著提高性能。在我们的最佳模型经过监督微调后，它在 CIFAR-10 数据集上，相对于排列不变方法中最低的错误率。