ClementPinard

import torch, kornia, h5py, imageio
import matplotlib.pyplot as plt
import numpy as np

# This is a failure mode, with different aspect ratios
fname1 = '47698078_3766965066'
fname2 = '18698491_4586522698'

# This works
# fname1 = '271147142_778c4e7999_o'

ClementPinard

from __future__ import division
from __future__ import print_function

import argparse
import math
import time
from tqdm import trange

import torch
from correlation_package.correlation import Correlation

ClementPinard

import time

import torch
import torch.backends.cudnn as cudnn
import torch.nn.functional as F

from torch.autograd import Variable

def benchmark_shape(N, C, IH, IW, H, W, nrand, nrep):
    """

ClementPinard

given $\theta_1$ and $\theta_2$ as affine transformations for $I_1$ and $I_2$ $$ \theta_1 = \left(\begin{array}{ccc}r_{0,1}&r_{1,1}&t_{0,1}\r_{2,1}&r_{3,1}&t_{1,1}\end{array} \right) = \left( R_1 | T_1 \right) $$ $$ \theta_2 = \left(\begin{array}{ccc}r_{0,2}&r_{1,2}&t_{0,2}\r_{2,2}&r_{3,2}&t_{1,2}\end{array} \right) = \left( R_2 | T_2 \right) $$

$$

ClementPinard

Still Box

Torrent description

In addition to this README, this torrent contains 4 datasets:

Name	Image size (px)	Scene number	Size compressed (B)	Total size (B)
64.tar.xz	64x64	80K	9.8G	19G
128.tar.xz	128x128	20K	7.1G	12G