skrish13

• Fine-Grained Few-Shot Classification with Feature Map Reconstruction Networks [Dec 2020]
• Variational Transfer Learning for Fine-grained Few-shot Visual Recognition [Oct 2020]
• Multi-attention Meta Learning for Few-shot Fine-grained Image Recognition [Jul 2020]
• Revisiting Pose-Normalization for Fine-Grained Few-Shot Recognition [Apr 2020]
• Compare More Nuanced-Pairwise Alignment Bilinear Network For Few-shot Fine-grained Learning [Apr 2019]
• Revisiting Local Descriptor based Image-to-Class Measure for Few-shot Learning [Mar 2019]
• Piecewise classifier mappings: Learning fine-grained learners for novel categories with few examples [May 2018]

skrish13

• RescueTime
• DueFocus
• Timecamp
• ManicTime [Privacy policy not updated since 2018]
• primeERP

skrish13

Price comparison

• https://mlcontests.com/cloud-gpu

General

• Google Cloud
• AWS
• Azure
• IBM AI
• Paperspace
• Lambdalabs

skrish13

SNo	Date	Topic
1	May 5th 2020	A Primer in BERTology
2	May 26th 2020	Dense Object Detection
3	June 3rd 2020	Neuroscience Part 1
4	June 10th 2020	Skinned Multi-Person Linear Model
5	June 25th 2020	Interpretable Models
6	July 8th 2020	Neuroscience Part 2

skrish13

def get_class(name):
    mod = __import__('folder.subfolder.{}'.format(name), fromlist=[''])
    return getattr(mod, _module_to_class(name))


def _module_to_class(name):
    return ''.join(n.capitalize() for n in name.split('_'))

skrish13

import torch as th                                                                 
                                                                                   
                                                                                   
class NLL_OHEM(th.nn.NLLLoss):                                                     
    """ Online hard example mining. 
    Needs input from nn.LogSotmax() """                                             
                                                                                   
    def __init__(self, ratio):      
        super(NLL_OHEM, self).__init__(None, True)                                 
        self.ratio = ratio                                                         

skrish13

def ROIAlign(feature_maps, rois, config, pool_size, mode='bilinear'):
    """Implements ROI Align on the features.
    Params:
    - pool_shape: [height, width] of the output pooled regions. Usually [7, 7]
    - image_shape: [height, width, chanells]. Shape of input image in pixels
    Inputs:
    - boxes: [batch, num_boxes, (x1, y1, x2, y2)] in normalized
             coordinates. Possibly padded with zeros if not enough
             boxes to fill the array.
    - Feature maps: List of feature maps from different levels of the pyramid.

skrish13

# Setup layers
self.mrcnn_mask_conv1 = nn.Conv2d(256, 256, kernel_size=3, stride=1, padding=1)
self.mrcnn_mask_bn1 = nn.BatchNorm2d(256, eps=0.001)
self.mrcnn_mask_conv2 = nn.Conv2d(256, 256, kernel_size=3, stride=1, padding=1)
self.mrcnn_mask_bn2 = nn.BatchNorm2d(256, eps=0.001)
self.mrcnn_mask_conv3 = nn.Conv2d(256, 256, kernel_size=3, stride=1, padding=1)
self.mrcnn_mask_bn3 = nn.BatchNorm2d(256, eps=0.001)
self.mrcnn_mask_conv4 = nn.Conv2d(256, 256, kernel_size=3, stride=1, padding=1)
self.mrcnn_mask_bn4 = nn.BatchNorm2d(256, eps=0.001)
self.mrcnn_mask_deconv = nn.ConvTranspose2d(256, 256, kernel_size=2, stride=2)

skrish13

wordlist created from original 41G stash via:

grep -rohP '(?<=:).*$' | uniq > breachcompilation.txt

Then, compressed with:

7z a breachcompilation.txt.7z breachcompilation.txt

Size:

skrish13

Dream Theater

• Self titled album
• Dramatic Turn of events

LTE

• http://www.geocities.ws/silkeryan/musicessay.htm