jenkspt

"""
This gist demonstrates the equivalence between the existing CLIP `AttentionPool2d` 
and the proposed `AttentionPool2dFix`, which only computes attention where needed.
"""

import torch
from torch import nn
import torch.nn.functional as F

jenkspt

import Base: zero, one


function uint_type(t)
    s = sizeof(t)
    if s == 2
        Int16
    elseif s == 4
        Int32
    elseif s == 8

jenkspt

def crop_center(image, crop_size):
    """ Crops the central region of the image with shape crop_size"""
    shape = image.shape

    start0 = shape[0]//2-(crop_size[0]//2)    
    start1 = shape[1]//2-(crop_size[1]//2)
    return img[start0:start0+crop_size[0],start1:start1+crop_size[1]]

jenkspt

import numpy as np
from sklearn.feature_extraction.image import extract_patches

def conv2d(inputs, filters):
    """
    Args:
        inputs (np.ndarray): NHWC
        filters (np.ndarray): 
            with shape [filter_height, filter_width, in_channels, out_channels]
    """

jenkspt

import random

def shuffle(iterator, buffer_size):
    """ Uses a buffer to randomly shuffle items from an iterator  """
    
    # Fill the buffer
    buffer = []; cnt = 0
    item = next(iterator, None)
    while not item is None and cnt < buffer_size:
        buffer.append(item)

jenkspt

import numpy as np
from numpy import linalg as la

"""
https://www.ltu.se/cms_fs/1.51590!/svd-fitting.pdf
"""

def get_rigid_transform(X, Y):
    """
    Calculates the rigid transform (translation, rotation)

jenkspt

import numpy as np
class Node(object):
    """
    Base class for nodes in the network.

    Arguments:

        `inbound_nodes`: A list of nodes with edges into this node.
    """
    def __init__(self, inbound_nodes=[]):

jenkspt

import numpy as np

def euclidean(A, B):
  return np.sqrt(((np.expand_dims(A, axis=-1) - B.T)**2).sum(axis=1))

jenkspt

import numpy as np
import scipy.stats as scs
import matplotlib.pyplot as plt

def plot(ax, x, y, label):
    lines = ax.plot(x, y, label=label, lw=2)
    ax.fill_between(x, 0, y, alpha=0.2, color=lines[0].get_c())

if __name__ == '__main__':
	# Simulated data for variations a and b