Discriminative Deep Metric Learning for Face Verification in the Wild

Hu, Junlin; Lu, Jiwen; Tan, Yap‐Peng

doi:10.1109/cvpr.2014.242

Cited by 607 publications

(377 citation statements)

References 33 publications

Supporting

Mentioning

370

Contrasting

Order By: Relevance

“…Given a test pair, we first extract the bag-level feature and then use the trained SVM to classify. For video-based face verification on the YTF dataset, we compare our method MILDIS with the several existing methods, including LM3L [18], DDML(LBP) [19], DDML(combined) [19], EigenPEP [20], DeepFace-single [3] and DeepID2+ [21]. The recognition accuracies and standard deviations of different methods are reported in Table 2.…”

Section: Experiments and Resultsmentioning

confidence: 99%

See 1 more Smart Citation

Multiple Instance Learning with Deep Instance Selection for Video-based Face Recognition

Liu¹

2016

Proceedings of the 2016 3rd International Conference on Mechatronics and Information Technology

View full text Add to dashboard Cite

Keywords: Multiple instance learning, deep learning, video-based face recognition. Abstract. In many real-world video-based face recognition scenarios, videos are usually captured under unconstrained conditions. It is very challenging because of low face resolutions, varying head pose and complex lighting. To address this issue, we present a new method by formulating the video-based face recognition issue as a multi-instance learning (MIL) problem. Specifically, given a pair of videos, we generate a bag composed of all the frame pairs from the two videos. The bag is positive if the given pair of videos is from the same person, otherwise it is negative. In this way, the recognition task is formulated as a binary classification problem in MIL. Then we propose a novel MIL algorithm with deep instance selection (MILDIS), which maps each bag into a feature space defined by the selected instances via an instance similarity measure. Our work achieves the state-of-the-art performances on the real-world datasets YouTube Faces (YTF) according to the restricted protocol.

show abstract

Section: Experiments and Resultsmentioning

confidence: 99%

“…From this table, out method achieves the state-of-the-art performances on the YTF dataset. [18] 81.3±1.2 DDML(LBP) [19] 81.3±1.6 DDML(combined) [19] 82.3±1.5 EigenPEP [20] 84.8±1.4 DeepFace-single [8] 91.4±1.1 DeepID2+ [21] 93.2±0.2 MILDIS 92.8±1.4…”

Section: Experiments and Resultsmentioning

confidence: 99%

Multiple Instance Learning with Deep Instance Selection for Video-based Face Recognition

Liu¹

2016

Proceedings of the 2016 3rd International Conference on Mechatronics and Information Technology

View full text Add to dashboard Cite

show abstract

“…Many approaches have been proposed in the literature that essentially exploit the label information from face images or face pairs. For instance, Hu et al [11] learn a discriminative metric within the deep neural network framework. Weinberger et al [9] propose Large Margin Nearest Neighbor (LMNN) metric which enforces the large margin constraint among all triplets of labeled training data.…”

Section: Introductionmentioning

confidence: 99%

An Efficient and Accurate Face Verification Method Based on CNN Cascade Architecture

Chen¹,

He²,

Yu³

et al. 2017

Proceedings of the 2017 2nd International Conference on Control, Automation and Artificial Intelligence (CAAI 2017)

View full text Add to dashboard Cite

Abstract-Unconstrained face verification has been actively studied for decades in computer vision. Recent algorithms rely on Convolution Neural Network to further improve the accuracy. However, such algorithms tend to be time-consuming and computationally complex, which cannot meet the real-time requirements. In this paper, we propose an efficient and accurate face verification method based on Convolution Neural Network Cascade architecture. First, we use a compact network to handle most of the simple samples. Then, we use a complex network to handle a small number of hard samples. Finally, we use an ensemble of multi-patch networks with metric learning. Our method achieves an accuracy of 99.72% on LFW, which performs favorably against the state-of-the-arts. Furthermore, we significantly reduce time cost from 485ms to only 20ms on a single core i7-4790, which has strong practical value for real-time face verification systems.

show abstract

“…For example, Yih et al [17] apply SFNNs to learn similarities on text data, Bordes et al [18] on entities in Knowledge Bases, and Masci et al [19] on multi-modal data. Other works on face verification use MLP Siamese architectures [20,21] or Restricted Boltzmann Machines (RBM) [22] to learn a non-linear similarity metric. 80 …”

mentioning

confidence: 99%

Class-balanced siamese neural networks

et al. 2018

View full text Add to dashboard Cite

This paper focuses on metric learning with Siamese Neural Networks (SNN). Without any prior, SNNs learn to compute a non-linear metric using only similarity and dissimilarity relationships between input data. Our SNN model proposes three contributions: a tuple-based architecture, an objective function with a norm regularisation and a polar sine-based angular reformulation for cosine dissimilarity learning. Applying our SNN model for Human Action Recognition (HAR) gives very competitive results using only one accelerometer or one motion capture point on the Multi-

show abstract

Discriminative Deep Metric Learning for Face Verification in the Wild

Cited by 607 publications

References 33 publications

Multiple Instance Learning with Deep Instance Selection for Video-based Face Recognition

Multiple Instance Learning with Deep Instance Selection for Video-based Face Recognition

An Efficient and Accurate Face Verification Method Based on CNN Cascade Architecture

Class-balanced siamese neural networks

Contact Info

Product

Resources

About