The Do’s and Don’ts for CNN-Based Face Verification

Bansal, Ankan; Castillo, Carlos D.; Ranjan, Rajeev; Chellappa, Rama

doi:10.1109/iccvw.2017.299

Cited by 72 publications

(46 citation statements)

References 45 publications

(183 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This suggests wider training datasets i.e. more identities, are indeed beneficial to CNN performance [6]. Experiment 2: Distractor Set.…”

Section: Experiments and Resultsmentioning

confidence: 97%

“…Our method can potentially create a synthetic distractor dataset, of any desired size, that can be used without running into potential issues of identity overlap or invasion of privacy. A dataset created with such synthetic face images will be free from any labeling errors like those that are found in public datasets such as VGG-Face [6,7].…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Fast Face Image Synthesis With Minimal Training

Banerjee

Scheirer

Bowyer

et al. 2019

2019 IEEE Winter Conference on Applications of Computer Vision (WACV)

View full text Add to dashboard Cite

We propose an algorithm to generate realistic face images of both real and synthetic identities (people who do not exist) with different facial yaw, shape and resolution. The synthesized images can be used to augment datasets to train CNNs or as massive distractor sets for biometric verification experiments without any privacy concerns. Additionally, law enforcement can make use of this technique to train forensic experts to recognize faces. Our method samples face components from a pool of multiple face images of real identities to generate the synthetic texture. Then, a real 3D head model compatible to the generated texture is used to render it under different facial yaw transformations. We perform multiple quantitative experiments to assess the effectiveness of our synthesis procedure in CNN training and its potential use to generate distractor face images. Additionally, we compare our method with popular GAN models in terms of visual quality and execution time.

show abstract

“…This suggests wider training datasets i.e. more identities, are indeed beneficial to CNN performance [6]. Experiment 2: Distractor Set.…”

Section: Experiments and Resultsmentioning

confidence: 97%

Section: Introductionmentioning

confidence: 99%

Fast Face Image Synthesis With Minimal Training

Banerjee

Scheirer

Bowyer

et al. 2019

2019 IEEE Winter Conference on Applications of Computer Vision (WACV)

View full text Add to dashboard Cite

show abstract

“…• We generalize one important finding of Bansal et al [15] to choose deeper or shallow architecture based on the depth of the dataset. In [15], they have reported the same in the context of face recognition, Whereas, we made a rigorous study to generalize this observation over different kinds of datasets.…”

Section: Necessity Of Fully Connected Layers In Cnnmentioning

confidence: 89%

“…Deeper vs Wider datasets [15]: For any two datasets with roughly same number of images, one dataset is said to be deeper [15]…”

Section: Crchistophenotypesmentioning

confidence: 99%

“…For example, the CNN-2 model produces the best accuracy of 92.29% for three FC layers with 4096, 256, and 10 neurons and the 2 nd best accuracy of 92.02% for two FC layers with 256 and 10 neurons.Bansal et al[15] have reported that the deeper architectures are preferred over shallow architectures while training the CNN models with deeper datasets.Whereas, for the wider datasets, the shallow architectures perform better compared to the deeper architectures. However, this observation is specific to face recognition problem as reported in[15]. In this paper, we made a rigorous study to generalize this finding by conducting extensive experiments on different modalities of datasets.…”

mentioning

confidence: 99%

See 1 more Smart Citation

Impact of fully connected layers on performance of convolutional neural networks for image classification

et al. 2020

View full text Add to dashboard Cite

The Convolutional Neural Networks (CNNs), in domains like computer vision, mostly reduced the need for handcrafted features due to its ability to learn the problem-specific features from the raw input data. However, the selection of dataset-specific CNN architecture, which mostly performed by either experience or expertise is a time-consuming and error-prone process. To automate the process of learning a CNN architecture, this paper attempts at finding the relationship between Fully Connected (FC) layers with some of the characteristics of the datasets. The CNN architectures, and recently datasets also, are categorized as deep, shallow, wide, etc. This paper tries to formalize these terms along with answering the following questions. (i) What is the impact of deeper/shallow architectures on the performance of the CNN w.r.t. FC layers?, (ii) How the deeper/wider datasets influence the performance of CNN w.r.t. FC layers?, and (iii) Which kind of architecture (deeper/shallower)is better suitable for which kind of (deeper/wider) datasets. To address these findings, we have performed experiments with three CNN architectures having different depths. The experiments are conducted by varying the number of FC layers. We used four widely used datasets including CIFAR-10, CIFAR-100, Tiny ImageNet, and CRCHistoPhenotypes to justify our findings in the context of image classification problem. The source code of this work is available at https://github.com/shabbeersh/Impact-of-FC-layers.

show abstract

Semi-supervised Adversarial Learning to Generate Photorealistic Face Images of New Identities from 3D Morphable Model

Geçer

Bhattarai

Kittler

et al. 2018

Computer Vision – ECCV 2018

View full text Add to dashboard Cite

We propose a novel end-to-end semi-supervised adversarial framework to generate photorealistic face images of new identities with wide ranges of expressions, poses, and illuminations conditioned by a 3D morphable model. Previous adversarial style-transfer methods either supervise their networks with large volume of paired data or use unpaired data with a highly under-constrained two-way generative framework in an unsupervised fashion. We introduce pairwise adversarial supervision to constrain two-way domain adaptation by a small number of paired real and synthetic images for training along with the large volume of unpaired data. Extensive qualitative and quantitative experiments are performed to validate our idea. Generated face images of new identities contain pose, lighting and expression diversity and qualitative results show that they are highly constraint by the synthetic input image while adding photorealism and retaining identity information. We combine face images generated by the proposed method with the real data set to train face recognition algorithms. We evaluated the model on two challenging data sets: LFW and IJB-A. We observe that the generated images from our framework consistently improves over the performance of deep face recognition network trained with Oxford VGG Face dataset and achieves comparable results to the state-of-the-art.

show abstract

The Do’s and Don’ts for CNN-Based Face Verification

Cited by 72 publications

References 45 publications

Fast Face Image Synthesis With Minimal Training

Fast Face Image Synthesis With Minimal Training

Impact of fully connected layers on performance of convolutional neural networks for image classification

Semi-supervised Adversarial Learning to Generate Photorealistic Face Images of New Identities from 3D Morphable Model

Contact Info

Product

Resources

About