BROWSE

Related Researcher

Author's Photo

Baek, Seungryul
Research Interests

ITEM VIEW & DOWNLOAD

Sampling Strategies for GAN Synthetic Data

DC Field Value Language
dc.contributor.author Bhattarai, Binod ko
dc.contributor.author Baek, Seungryul ko
dc.contributor.author Bodur, Rumeysa ko
dc.contributor.author Kim, Tae-Kyun ko
dc.date.available 2020-06-24T00:56:39Z -
dc.date.created 2020-04-21 ko
dc.date.issued 2020-05-04 ko
dc.identifier.citation IEEE International Conference on Acoustics, Speech and Signal Processing ko
dc.identifier.uri https://scholarworks.unist.ac.kr/handle/201301/32633 -
dc.description.abstract Generative Adversarial Networks (GANs) have been used widely to generate large volumes of synthetic data. This data is being utilized for augmenting with real examples in order to train deep Convolutional Neural Networks (CNNs). Studies have shown that the generated examples lack sufficient realism to train deep CNNs and are poor in diversity. Unlike previous studies of randomly augmenting the synthetic data with real data, we present our simple, effective and easy to implement synthetic data sampling methods to train deep CNNs more efficiently and accurately. To this end, we propose to maximally utilize the parameters learned during training of the GAN itself. These include discriminator's realism confidence score and the confidence on the target label of the synthetic data. In addition to this, we explore reinforcement learning (RL) to automatically search a subset of meaningful synthetic examples from a large pool of GAN synthetic data. We evaluate our method on two challenging face attribute classification data sets viz. AffectNet and CelebA. Our extensive experiments clearly demonstrate the need of sampling synthetic data before augmentation, which also improves the performance of one of the state-of-the-art deep CNNs in vitro. ko
dc.publisher Institute of Electrical and Electronics Engineers Inc. ko
dc.title Sampling Strategies for GAN Synthetic Data ko
dc.type CONFERENCE ko
dc.type.rims CONF ko
dc.identifier.url https://arxiv.org/abs/1909.04689 ko
Appears in Collections:
AI_Conference Papers

find_unist can give you direct access to the published full text of this article. (UNISTARs only)

Show simple item record

qrcode

  • mendeley

    citeulike

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.

MENU