발표 순서 재조정

김승일, 이일구, 백병인, 홍기호, 한성국, 이동훈, 김정석, 박두호, 최병국, 현청천, 최민영, 정규민, 장주영, 이혜민

====

정자민, 전인태, 김혜린, 박상진, 구태훈, 고병건, 박철, 이용근, 김일승, 김종목, 하종수, 신현철

====

 

4월 8일 : 조원양, 이일구

High Performance Visual Tracking with Siamese Region Proposal Network

Exploring Randomly Wired Neural Networks for Image Recognition

 

4월 1일 : 한성국, 홍기호

BERT has a Mouth, and It Must Speak: BERT as a Markov Random Field Language Model

Do Better ImageNet Models Transfer Better?

 

3월 25일 : 백병인, 이동훈

GPT-2를 비롯하여 그동안 NLP 연구 흐름

Deep Lagrangian Networks: Using Physics as Model Prior for Deep Learning

 

 

3월 18일 : 이혜민, 이동열

Adversarial Autoencoders

Bayesian Convolutional Neural Networks with Variational Inference

 

3월 11일 : 김정석, 장주영

Automated Breast Ultrasound Lesions Detection Using Convolutional Neural Networks

Convolutional CRFs for Semantic Segmentation

 

3월 4일 : 이일구, 홍기호

Invertible Residual Networks

Bag of Tricks for Image Classification with Convolutional Neural Networks

 

2월 25일 : 한성국, 이동훈

Building Dynamic Knowledge Graphs from Text using Machine Reading Comprehension
WaveGlow

 

2월 18일 : 백병인, 김승일

Dendritic cortical microcircuits approximate the backpropagation algorithm
Audio Synthesis using Generative Models

 

2월 11일 : 장주영, 이혜민
Video-to-Video Synthesis
Pelee: A Real-Time Object Detection System on Mobile Devices

2월 4일 : 설연휴

 

1월 28일 : 최병국, 김정석

AtomNet: A Deep Convolutional Neural Network for Bioactivity Prediction in Structure-based Drug Discovery
GAN-based Synthetic Medical Image Augmentation for increased CNN Performance in Liver Lesion Classification

 

1월 21일 : 이일구, 현청천

How Does Batch Normalization Help Optimization? (No, It Is Not About Internal Covariate Shift)

Emotional Chatting Machine: Emotional Conversation Generation with Internal and External Memory

 

1월 14일 : 이동훈, 홍기호

Glow: Generative Flow with Invertible 1x1 Convolutions

Domain Adaptive Transfer Learning with Specialist Models

 

1월 7일 : 김승일, 한성국

BERT : Pre-training of Deep Bidirectional Transformer for Language Understanding

Universal Transformers

 

======== 2019년 ========

 

12월 3일 : 이일구, 백병인

An overview of gradient descent optimization algorithms

Bottom-Up and Top-Down Attention for Image Captioning and Visual Question Answering

 

11월 26일 : 박두호, 현청천

DeepLog: Anomaly Detection and Diagnosis from System Logs through Deep Learning

Efficient and Robust Question Answering from Minimal Context over Documents

 

11월 19일 : 김정석, 최병국

UNet++: A Nested U-Net Architecture for Medical Image Segmentation

DroNet: Learning to Fly by Driving

 

11월 12일 : 이일구, 홍기호

NICE: Non-linear Independent Components Estimation

Wide Residual Networks

 

11월 5일 : 한성국, 이동훈

QANet: Combining Local Convolution with Global Self-Attention for Reading Comprehension

Spherical CNNs

 

10월 29일 : 현청천, 이일구

Deep contextualized word representations

Wave-U-Net: A Multi-Scale Neural Network for End-to-End Audio Source Separation

 

10월 22일 : 김승일, 백병인

lip reading sentences in the wild

Self-Attention with Relative Position Representations

 

10월 8일 : 박두호, 이동훈

Large Scale GAN Training for High Fidelity Natural Image Synthesis

Can recurrent neural networks warp time?

 

10월 1일 : 한성국, 최병국

Sequence Modelling with Memory-Augmented Recurrent Neural Networks

A Unified Multi-scale Deep Convolutional Neural Network for Fast Object Detection

 

## Attention 관련 논문 시리즈

9월 17일 : 홍기호, 김정석

CBAM: Convolutional Block Attention Module

Attention-Gated Networks for Improving Ultrasound Scan Plane Detection

 

9월 10일 : 장주영, 백병인

A Decomposable Attention Model for Natural Language Inference

Strunctured Attention Network (백병인)
Attention is all you need (백병인)

 

9월 3일 : 김은지, 이동훈

Neural machine translation by jointly learning to align and translate

Show attend and tell

## Attention 관련 논문 시리즈

 

8월 27일 : 박두호, 김일승

Training Deep Networks with Synthetic Data: Bridging the Reality Gap by Domain Randomization

ShuffleNet V2

 

8월 20일 : 한성국, 이동훈

Jointly Trained Sequential Labeling and Classification by Sparse Attention Neural Networks

이동훈님 두개 요약발표

Style Tokens: Unsupervised Style Modeling, Control and Transfer in End-to-End Speech Synthesis

Transfer Learning from Speaker Verification to Multispeaker Text-To-Speech Synthesis

 

8월 13일 : 김일승, 이일구

Context-aware Deep Feature Compression for High-speed Visual Tracking

Demystifying Neural Style Transfer

 

8월 6일 : 홍기호

Why do deep convolutional networks generalize so poorly to small image transformations?

 

7월 30일 : 휴가

 

7/23 : 백병인, 김정석

Model-based Reinforcement Learning Survey

 

## Generative Adversarial Network 시리즈 

7/16 김혜린,

Domain Adaptation, Domain transfer using GAN

 

7/9 김승일, 박두호, 장주영

Self-Attention Generative Adversarial Networks

Progressive Growing of GANs

StarGAN

 

7/2 이일구, 이동훈

Generative Adversarial Network Intro (gan, dcgan, cgan, infogan, bigan, wgan, began, pix2pix, cyclegan)
Generative Adversarial Network Intro (lsgan, fgan, unrolledgan, wgan)

## Generative Adversarial Network 시리즈 

 

6/25 진영재, 한성국

 

6/18 정자민, 강병익

MobileNetV2: Inverted Residuals and Linear Bottlenecks

 

6/11 조익연, 이정은

 

6/4 고병건, 박상진

 

Convolutional Neural Networks Applied to High-Frequency Market Microstructure Forecasting

 

5/28 곽대훈, 이동훈

Born Again Neural Networks

Learning Beyond Human Expertise with Generative Models for Dental Restorations

 

5/21 구태훈, 전인태

Prioritized experience replay

 

5/14 이상헌, 홍기호

Complex-YOLO: Real-time 3D Object Detection on Point Clouds

Large-Scale Image Retrieval with Attentive Deep Local Features

 

4/30 이일구, 백병인

Thinking Fast and Slow with Deep Learning and Tree Search

A Closed-form Solution to Photorealistic Image Stylization

 

4/23 김승일, 이동훈

Tacotron 2

Deep Voice 2, 3

 

4/16 김승일, 이동훈

Tacotron : Towards End-to-End Speech Synthesis

Deep Voice

 

4/9 김영주, 김혜린

Neural Machine Translation by Jointly Learning to Align and Translate

Visual attribute transfer through deep image analogy

 

4/2 한성국, 곽대훈

Poincare Embeddings for Learning Hierarchical Representations

Lifelong Learning with Dynamically Expandable Networks

 

3/26 김정석, 진영재

Deep Learning in RF Sub-sampled B mode Ultrasound Imaging

 

3/19 정자민, 강병익

MarrNet: 3D Shape Reconstruction via 2.5D Sketches

Modelling the Scene Dependent Imaging in Cameras with a Deep Neural Networks (ICCV 2017)

 

3/12 이동석, 김지훈

Squeeze-and-Excitation Networks

Spherical CNNs

 

3/5 이동훈, 홍기호

Q-Learning-based financial trading systems with applications

Deep Metric Learning via Facility Location

 

2/26 전인태, 이상헌

Blockchain and Artificial Intelligence (II)

Multi-View 3D Object Detection Network for Autonomous Driving

 

2/19 전인태, 한성국

Blockchain and Artificial Intelligence (I)

The Numerics of GANs

 

2/12 곽대훈, 김승일

Gradient Episodic Memory for Continual Learning

Deep Learning vs Professional Healthcare Equipment: A Fine-Grained Breathing Rate Monitoring Model

 

2/5 전체회식

 

1/29 강병익, 이일구

BlitzNet: A Real-Time Deep Network for Scene Understanding

Visualizing the Loss Landsacpe of Neural Nets

 

1/22 김정석, 정자민

Interpretable Deep Models for ICU Outcome Prediction

BoxCars: Improving Fine-Grained Recognition of Vehicles using 3D Bounding Boxes in Traffic Surveillance

 

1/15 김지훈, 이상헌

Outrageously Large Neural Networks: The Sparsely-Gated Mixture-of-Experts Layer

PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation (Part II)

 

1/8 김승일, 이상헌

Connecting Generative Adversarial Networks and Actor-Critic Methods

PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation (Part I)

 

======== 2018년 ========

12/25 모두연 송년파티 (예정)

 

12/18 한성국, 이동훈

A Hybrid Convolutional Variational Autoencoder for Text Generation

Matrix Capsules with EM Routing

 

12/11 이영한, 홍기호

Deep Voice 3: 2000-Speaker Neural Text-to-Speech

Building Mobile Applications of TensorFlow

 

12/4 강병익, 이일구

Enhanced Deep Residual Networks for Single Image Super-Resolution

Generative Adversarial Networks: An Overview

 

11/27 박상진, 정자민

Capsule Net

Spatio-Temporal LSTM with Trust Gates for 3D Human Action Recognition

 

11/20 양종렬, 손준영

Deep Learning for Event-Driven Stcok Prediction

Image Captioning and Classification of Dangerous Situations

 

11/3 김지훈, 이상헌

Towards Principled Methods for Training Generative Adversarial Networks

Predicting the Driver's Focus of Attention: the DR(eye)VE Project

 

11/6 이영한, 김정석

Look, Listen and Learn

Spectral Graph Convolutions for Population-based Disease Prediction

 

10/30 고민수, 홍기호

Learning by Association A versatile semi-supervised training method for neural networks

Learning From Noisy Large-Scale Datasets With Minimal Supervision

 

10/23 박은수, 김승일

InfoGAN: Interpretable Representation Learning by Information Maximizing Generative Adversarial Nets

A Tutorial on Deep Learning for Music Information Retrieval

 

10/16 이동훈(2부), 이동석

EraseReLU : A SIMPLE WAY TO EASE THE TRAINING OF DEEP CONVOLUTION NEURAL NETWORKS

계속

 

9/25 박상진, 이동훈

Faster R-CNN

Fast Gradient-Descent Methods for Temporal-Difference Learning with Linear Function Approximation

 

9/18 한영일, 송준호

Multi-view Self-supervised Deep Learning for 6D Pose Estimation in the Amazon Picking Challenge

Understanding deep learning requires rethinking generalization

 

9/11 이일구, 조익연

How does physics connect to machine learning?

Seq2SQL: Generating Structured Queries from Natural Language using Reinforcement Learning

 

9/4 김정석, 김지훈 part 2

Doctor AI: Predicting Clinical Events via Recurrent Neural Networks

 

8/28, 이상헌, 김지훈

End-to-end Learning of Driving Models from Large-scale Video Datasets

Understanding Black-box Predictions via Influence Functions

 

8/21 조용래, 이영한

A SIMPLE BUT TOUGH-TO-BEAT BASELINE FOR SENTENCE EMBEDDINGS

Delving into Transferable Adversarial Examples and Black-box Attacks

 

8/14 이동훈, 김선훈, 한성국 서기호, 최윤규 

One-shot imitation learning

VNect: Real-time 3D Human Pose Estimation with a Single RGB Camera

 

8/7 홍기호, 고민수

Beyond instance-level image retrieval: Leveraging human captions to learn representations for semantic visual search

Look Closer to See Better: Recurrent Attention Convolutional Neural Network for Fine-grained Image Recognition

 

7/31 이용근, 김승일

Densely Connected Convolutional Networks

Multi-Agent Actor-Critic for Mixed Cooperative-Competitive Environments

 

7/24 고정섭, 박상진

In-Datacenter Performance Analysis of a Tensor Processing Unit​

Hierarchical Deep Reinforcement Learning: Integrating Temporal Abstraction and Intrinsic Motivation

 

7/17 한영일, 송준호

Very Deep Convolutional Networks for Text Classification

Revisiting Unreasonable Effectiveness of Data in Deep Learning Era

 

7/10 백병인, 조익연

Attention is all you need

One Model to Learn Them All

 

7/3 이영한, 김정석님

Tacotron: Towards End-to-End Speech Synthesis

Computer-Aided Diagnosis with Deep Learning Architecture: Applications to Breast Lesions in US Images and Pulmonary Nodules in CT Scans

 

6/26 김지훈님

Realtime Multi-Person 2D Pose Estimation using Part Affinity Fields

 

6/19

이일구, 조용래님

Self-Nomalizing Neural Networks

A simple neural network module for relational reasoning

 

6/12

서기호, 이상헌님

Neural Architecture Search with Reinforcement Learning

Brain4Cars: Car that knows before you do via Sensory-Fusion Deep Learning Architecture

 

6/5

한성국, 김선훈님

Bayesian Recurrent Neural Networks

Controllable Text Generation

 

5/29

백병인, 고민수님

Continual Learning with Deep Generative Replay

Understanding Deep Learning Requires Rethinking Generalization

 

5/22

김승일, 홍기호님

Continuous Control with Deep Reinforcment Learning

Convolutional Sequence to Sequence Learning

 

5/15

이영한, 김정석님

Distilling the Knowledge in a Neural Network

Interpretable Deep Models for ICU Outcome Prediction

 

5/8

홍순원님, 서기호님

Neural audio synthesis of musical notes with WaveNet autoencoders

Safe, Multi-Agent, Reinforcement Learning for Autonomous Driving

 

4/24 - 이일구, 한영일님

Auto-Encoding Variational Bayes

Tutorial on Variational Autoencoders

Show and Tell: A Neural Image Caption Generator

Show, Attend and Tell : Neural Image Caption Generation with Visual Attention

DenseCap: Fully Convolutional Localization Networks for Dense Captioning

 

4/17 - 고민수, 한성국님

Linguistic Knowledge as Memory for Recurrent Neural Networks : https://arxiv.org/pdf/1703.02620.pdf

 

Mask R-CNN https://arxiv.org/abs/1703.06870

 

4/10 - 김승일, 홍기호님

SEGAN : Speech Enhancement Generative Adversarial Network

논문링크 : https://arxiv.org/abs/1703.09452

NIPS2016 Tutorial : Variational Inference 2부

링크 : https://nips.cc/Conferences/2016/Schedule?type=Tutorial

 

4/3 - 백병인, 이상헌님

DRAGNN - A Transition-based Framework for Dynamically Connected Neural Network

논문 링크 : https://arxiv.org/abs/1703.04474

Learning Spatiotemporal Features with 3D Convolutional Networks

논문 링크 : https://arxiv.org/abs/1412.0767

 

3/27 - 김태완, 이일구님

YOLO9000 : Better, Faster, Stronger

https://arxiv.org/abs/1612.08242

Learning to Discover Cross-Domain Relations with Generative Adversarial Networks (DiscoGAN)

https://arxiv.org/pdf/1703.05192.pdf

 

 

3/20 - 성태경, 이용근님

LipNet : End-to-End Sentence-Level Lipreading https://arxiv.org/abs/1611.01599

SqueezeDet: Unified, Small, Low Power Fully Convolutional Neural Networks for Real-Time Object Detection for Autonomous Driving https://arxiv.org/abs/1612.01051

 

3/13 - 홍순원님, 이영한님

PathNet : Evolution Channels Gradient Descent in Super Neural Networks

Matching Networks for One shot learning

 

3/6 - 고민수님, 김정석님

Image-to-Image Translation with Conditional Adversarial Networks

Learning the Number of Neurons in Deep Networks

 

2/27일 - 한성국님

NIPS Tutorial : Large-Scale Optimization: Beyond Stochastic Gradient Descent and Convexity

 

2/20일 - 김승일님

NIPS Tutorial : Deep Reinforcment Learnining through Policy Optimization

 

2/13일 - 홍기호님

NIPS Tutorial : Variational Inference - Foundations and Modern Methods

 

2/6일 - 이일구님

NIPS Tutorial : Generative Adversarial Networks (Ian Goodfellow)

 

1/23일 - 이용근님, 김승일님

Identity Mappings in Deep Residual Network

One-Shot Learning with Memory-Augmented Neural Networks

 

1/16일 - 김정석님, 한성국님

Deep Patient: An Unsupervised Representation to Predict the Future of Patients from the Electronic Health Records

Recurrent Highway Networks

 

1/9일 - 이영한님 , 고민수님

Phased LSTM

Binarized Neural Networks: Training Deep Neural Networks with Weights and Activations Constrained to +1 or -1

 

1/2일 -  홍기호님, 서기호님

Swivel: Improving Embeddings by Noticing What's Missing

Learning to Compose Neural Networks for Question Answering

NeuralModuleNetworks

 

=====

기존의 발표논문 리스트 (~2016.12)

- SampleRNN : An Unconditional End-to-End Neural Audio Generation Model

- Swapout: Learning an ensemble of deep architectures

- Adversarially Learned Inference

- SqueezeNet: AlexNet-level accuracy with 50x fewer parameters and <0.5MB model size  (revisited)

- Learning Fine-grained Image Similarity with Deep Ranking

- Particular object retrieval with integral max-pooling of CNN activations

- Generative Adversarial Text to Image Synthesis

- Real-Time Single Image and Video Super-Resolution Using an Efficient Sub-Pixel Convolutional Neural Network

- Asynchronous Methods for Deep Reinforcement Learning

- Deep Reinforcement Learning with Double Q-Learning

- Learning Deep Features for Discriminative Localization

- Google's Neural Machine Translation System: Bridging the Gap between Human and Machine Translation

- Visual Fashion-Product Search at SK Planet

- A Machine Learning Approach to Visual Perception of Forest Trails for Mobile Robots

- Deep Neural Network for Real-Time Autonomous Indoor Navigation

- Deep Compression: Compressing Deep Neural Networks with Pruning, Trained Quantization and Huffman Coding

- Using Fast Weights to Attend to the Recent Past

- Learning both Weights and Connections for Efficient Neural Networks

- hybrid Computing using a neural network with dynamic memory

- Reward augmented maximum likelihood for neural structured prediction

- Intriguing properties of neural networks

- A Fully Convolutional Neural Network for Speech Enhancement

- Human-level control through deep reinforcement learning (DQN)

- Decoupled Neural Interfaces using Synthetic Gradients

- Wavenet

- Q-Learning

- Deep Speech 2

- Image Segmentation 총정리 :

- Semantic Image Segmentation with Task-Specific Edge Detection Using CNNs and a Discriminatively Trained Domain Transform

- Multi-view Face Detection using Deep Convolutional Neural Networks

- A Convolutional Neural Network Cascade for Face Detection

- Inside-Outside Net : Detecting Objct in Context with Skip Pooling and Recurrent Neural Networks

- TensorFlow: A system for large-scale machine learning

- Perceptual Losses for Real-Time Style Transfer and Super-Resolution

- Inception Trilogy 1,2,3편

- OpenFace

- Learning to Respond with Deep Neural Networks for Retrieval-Based Human-Computer Conversation System

- Deep Speech: Scaling up end-to-end speech recognition

- SSD : Single Shot MultiBox Detector

- A Fast and Accurate Dependency Parser using Neural Networks

- Adieu Features? End-to-End Speech Emotion Recognition using a Deep Convolutoinal Recurrent Network

- Deep Learning for Robust Feature Generation in AudioVisual Emotion Recognition

- Video Emotion Recognition with Transferred Deep Feature Encodings

- Deep Network with Stochastic Depth

- Towards Good Practices for Very Deep Two-Stream ConvNets

- Deep Metric Learning using TripleNet Network

- Delving Deeper into Convolutional Networks for Learning Video Representations

- Learning Spatiotemporal Features with 3D Convolutional Networks

- Multi-Scale Context Aggregation by Dialated Convolutions

- Colorful Image Colorization

- SqeezeNet : AlexNet-Level Accuracy with 50x Fewer Parameters and 1MB Model Size

- Word2Vec

- U-Net : Convolutional Networks for Biomedical Image Segmentation

- Exploring the Neural Algorith of Artistic Style

- VGGNet and ResNet

- Recurrent Memory Network for Language Modeling

- Grid Long Short-Term Memory

- Learning Deconvolution Network for Semantic Segmentation

- Fully Convolutional Networks for Semantic Segmentation

- You Only Look Once - Unified Real Time Object Detection

- Texture Synthesis using Convolutional Neural Networks

- Unsupervised Representation Learning with Deep Convolution Generative Adversarial Networks

- Spatial Transformer Networks

- Generating Images from Captions with Attention

- Generating Sequences with Recurrent Neural Networks

- End-to-End Memory Networks

- AlexNet

- Overfeat : Integrated Recognition, Localization and Detection using Convolutional Networks

- Recommender System using Deep Learning

- Restricted Boltzmann Machines for Collaborative Filtering

- Collaborative Deep Learning for Recommender Systems

- Deep Content-Based Music Recommendataion

- A Neural Algorithm of Artistic Style

- A Unified Framework for Automatic Wound Segmentation and Analysis with Deep Convolutional Neural Network