Home

RoIPooling Explanation Region of interest pooling (also known as RoI pooling) is an operation widely used in object detection tasks using convolutional neural networks. For example, to detect multiple cars and pedestrians in a single image. Its purpose is to perform max pooling on inputs of nonuniform sizes to obtain fixed-size feature maps (e....

Info Title: InstaGAN: Instance-aware Image-to-Image Translation Task: Image-to-Image Translation Author: Sangwoo Mo, Minsu Cho, Jinwoo Shin Date: Dec. 2018 Arxiv: 1812.10889 Published: ICLR 2019 Highlights & Drawbacks Instance-level translation with semantic map Sequential mini-batch training strategy Abstract Unsuperv...

ProGAN ProGAN is a new technique developed by NVIDIA Labs to improve both the speed and stability of GAN training.6 Instead of immediately training a GAN on full-resolution images, the paper suggests first training the generator and discriminator on low-resolution images of, say, 4 × 4 pixels and then incrementally adding layers throughout the t...

Project page Paper GTC 2019 demo Youtube Demo of GauGAN Installation Clone this repo. git clone https://github.com/NVlabs/SPADE.git cd SPADE/ This code requires PyTorch 1.0 and python 3+. Please install dependencies by pip install -r requirements.txt This code also requires the Synchronized-BatchNorm-PyTorch rep....

Info Title: Point-to-Point Video Generation Task: Video Generation Author: Tsun-Hsuan Wang, Yen-Chi Cheng, Chieh Hubert Lin, Hwann-Tzong Chen, Min Sun Date: Apr. 2019 Arxiv: 1904.02912 Published: ICCV 2019 Abstract While image manipulation achieves tremendous breakthroughs (e.g., generating realistic faces) in recent years, video...

Info Title: MoCoGAN: Decomposing Motion and Content for Video Generation Task: Video Generation Author: Sergey Tulyakov, Ming-Yu Liu, Xiaodong Yang, Jan Kautz Date: July 2017 Arxiv: 1707.04993 Published: CVPR 2018 Abstract Visual signals in a video can be divided into content and motion. While content specifies which objects are ...

PSNR: Peak Signal-to-Noise Ratio Peak signal-to-noise ratio, often abbreviated PSNR, is an engineering term for the ratio between the maximum possible power of a signal and the power of corrupting noise that affects the fidelity of its representation. The Formula where Numpy Implementation import math import numpy as np def calculate_ps...

The following script down samples HR images to LR images for Super-Resolution data preparation. It requires OpenCV and NumPy. Python Version import os import sys import cv2 import numpy as np try: sys.path.append(os.path.dirname(os.path.dirname(os.path.abspath(__file__)))) from data.util import imresize_np except ImportError: pass...