DeepLearning(11)
-
[Deep Learning] Conv1D에 관하여
CNN은 일반적으로 이미지에서 계층적 특징 추출을 위해 사용된다. CNN의 이러한 장점을 활용하여 2차원 이미지가 아닌 1차원의 sequential 데이터에도 CNN이 사용된다. 주어진 sequence data에서 중요한 정보를 추출해낼 수 있다. 1D filter shape: [height, n] n: input data embedding dim (fixed value) filter size에서 변경 가능한 값은 height filter는 수직 방향으로만 움직인다. NLP에서는 height는 몇 개 단어를 고려할 지 결정하는 값 Pros and Cons Pros 시간에 따라 기록된 센서 데이터를 처리하는 데 용이하다 (Ex. audio signals) 고정된 길이의 데이터를 처리하는 데 용이하다 "Sp..
2020.09.29 -
[Paper Review] RandAugment: Practical automated data augmentation with a reduced search space
Overview 제목: RandAugment: Practical automated data augmentation with a reduced search space 저자: Ekin D. Cubuk , Barret Zoph , Jonathon Shlens, Quoc V. Le 기관: Google Brain Paper: https://arxiv.org/abs/1909.13719 Summary: automated data augmentation 기법으로써, 오직 두 개의 parameter (N,M) 를 이용하여 기존 Automated Augmentation 기법보다 매우 간단한데 더 좋은 성능을 냄. 기존 연구 기존 automated augmentation 기법(AutoAugment, Fast Autoaugm..
2020.08.28 -
[Paper Review] GANSpace: Discovering Interpretable GAN Controls
Overview 제목: GANSpace: Discovering Interpretable GAN Controls 저자: Erik Härkönen, Aaron Hertzmann, Jaakko Lehtinen, Sylvain Paris 기관: Aalto Univ., Adobe Research, NVIDIA 학회: ECCV 2020 under review 요약: "Unsupervised identification of interpretable directions in an existing GAN" Pre-trained StyleGAN, BigGAN의 초반 activation space에서 PCA를 수행하여 얻은 각 component를 조절하여 image manipulation을 수행 PCA로 GAN의 laten..
2020.08.27 -
[Paper Review] Expert-level detection of acute intracranial hemorrhage on head computed tomography using deep learning
Title: Expert-level detection of acute intracranial hemorrhage on head computed tomography using deep learning Conference: PNAS Summary: Architecture: PatchFCN, a patch-based fully convolutional neural network. Being a workhorse in medical imaging modality, Brain Computed Tomography(CT) needs a deep learning that can accurately identify diverse and very subtle cases of a major class of pathology..
2020.08.19 -
[Paper Review] The Unreasonable Effectiveness of Deep Features as a Perceptual Metric
Title: The Unreasonable Effectiveness of Deep Features as a Perceptual Metric Authors: Richard Zhang(1), Phillip Isola(1,2), Alexei A. Efros(1), Eli Shechtman(3), Oliver Wang(3) Affiliation: UC Berkeley(1), OpenAI(2), Adobe Research(3) Official Code: Link Summary PROBLEM STATEMENT: Evaluate how well metrics correspond with human perceptual judgments. Collect a large-scale perceptual similarity d..
2020.08.13 -
[Paper Review]Skip-GANomaly
Skip-GANomaly: Skip Connected and Adversarially Trained Encoder-Decoder Anomaly Detection, 2019. Authors: Samet Akçay, Amir Atapour-Abarghouei, Toby P. Breckon Conference: 2019 International Joint Conference on Neural Networks (IJCNN) Github: Link Summary: an U-net like encoder-decoder convolutional neural network with skip-connections and utilized adversarial training scheme the role of skip co..
2020.08.12