myungsub

---
description: 
globs: 
alwaysApply: true
---

## Core Directive
You are a senior software engineer AI assistant. For EVERY task request, you MUST follow the three-phase process below in exact order. Each phase must be completed with expert-level precision and detail.

## Guiding Principles

myungsub

• Improving Pairwise Ranking for Multi-label Image Classification
• Improved Deep Metric Learning with Multi-class N-pair Loss Objective
• Dimensionality Reduction by Learning an Invariant Mapping
• Exponential discriminative metric embedding in deep learning
• Learning Universal Embeddings from Attributes
• TVAE: TRIPLET-BASED VARIATIONAL AUTOENCODER USING METRIC LEARNING
• [Simple Triplet Loss Based on Intra/Inter-class Metric Learning for Face Verification](http://openaccess.thecvf.com/content_ICCV_2017_workshops/papers/w23/Ming

myungsub

Papers from Super SloMo references

• Super SloMo: High Quality Estimation of Multiple Intermediate Frames for Video Interpolation [Paper]
  • Huaizu Jiang, Deqing Sun, Varun Jampani, Ming-Hsuan Yang, Erik Learned-Miller, Jan Kautz
  • CVPR 2018 (splotlight)
• Video frame synthesis using deep voxel flow [Paper] [Code]
  • Z. Liu, R. Yeh, X. Tang, Y. Liu, and A. Agarwala.
  • ICCV 2017
• Video frame interpolation via adaptive separable convolution. [Paper] [Code]

myungsub

How I gather all the (research-related) information

1. Twitter
2. Facebook
3. Reddit - Machine Learning subreddit
4. Arxiv-Sanity
5. Github
6. Conference accepted papers list
7. Conference workshop accepted papers list
8. Manual wandering of famous blogs

myungsub

NIPS 2017

• Temporal Coherency based Criteria for Predicting Video Frames using Deep Multi-stage Generative Adversarial Networks
• Prototypical Networks for Few-shot Learning
• Good Semi-supervised Learning That Requires a Bad GAN
• Neural Discrete Representation Learning
• Learning Disentangled Representations with Semi-Supervised Deep Generative Models

myungsub

• Self-Transfer Learning for Fully Weakly Supervised Object Localization
  • by Sangheum Hwang and Hyo-Eun Kim (both from Lunit)
  • Arxiv:1602, MICCAI 2016
  • Datasets used: KIT(training), Shenzhen(test), MC set(test)
    • Localization annotation is obtained by human clinicians
• LEARNING TO DIAGNOSE FROM SCRATCH BY EXPLOITING DEPENDENCIES AMONG LABELS [paper]
  • Li Yao, Eric Poblenz, Dmitry Dagunts, Ben Covington, Devon Bernard, Kevin Lyman
  • Arxiv:1710
• Learning to detect chest radiographs containing lung nodules using visual attention networks [paper]
• Emanuele Pesce, Petros-Pavlos Ypsilantis, Samuel Withey, Robert Bakewell, Vicky Goh, Giovanni Montana

myungsub

Github with Mathjax Chrome Extension Test

$$ E = mc^2 $$ I want this equation $ \alpha = \beta + \gamma $ to render correctly

myungsub

{"lastUpload":"2017-04-10T12:47:54.193Z","extensionVersion":"v2.6.2"}

myungsub

# Jupyter Notebook Tutorial
https://www.digitalocean.com/community/tutorials/how-to-set-up-a-jupyter-notebook-to-run-ipython-on-ubuntu-16-04

myungsub

cjson = require 'cjson'
f = io.open(filename, 'r')
text = f:read()
info = cjson.decode(text)

-- 'info' is a table containing everything
-- info.all_losses
-- info.iter
-- info.loss_history
-- info.results_history