Image classification frescoplay solutions

An input image can be converted into structured form through ___________
Image Analysis

CNN is a algorithm in which it comes under
Deep Learning

Confusion Matrix is not a technique used to evaluate the performance of a classifier
FALSE

Conventional classification algorithms on image data
could not give significant accuracy

Identify the unstructured data from the following.
Both image and video clip

The major steps involved in image classification are ___________
Input Image -> Preprocessing -> Feature Extraction -> Classification

Noise can be removed using Data Preprocessing
TRUE

The data is split into _________ sets
3

The major function of PCA is to decompose a ________________________ into a set of successive orthogonal components
Multivariate dataset

Cross validation gives high variance if the testing set and training set are not drawn from the ______________________
Same Population

A normalized image has ___________
Mean = 0 and variance = 1

SIFT is mainly used for images that are ___________
less simpled and less organised

SVD is used in many fields
TRUE

For SVM it is good to have ___________
number of dimensions > the number of samples

Normalization is the process of converting ___________
pixel values to normal state

The main aim of using SIFT for feature extraction is to obtain features that are very sensitive to changes in scale, rotation, image resolution, illumination, etc.
FALSE

The scale-invariant feature transform can be used to detect and describe local features in images.
TRUE

Each class in CIFAR -10 dataset has _________ images.
6000

The data used to tune the model is _____________
Validation Set

The number of incorrect predictions that the occurrence is negative is False Negative
TRUE

___________________________ is defined as the percentage of correct predictions
Classification Accuracy


Naive Bayes algorithm comes under
Deep Learning

In the SVD method, a digital image is decomposed into ________ matrices
3

The process of changing the pixel intensity values to achieve consistency in dynamic range for images is ___________.
Image normalisation

The steps involved in building a classification model are -
Initialize---->Train---->Predict---> Evaluate

Data Preprocessing is a step the raw data is converted into a form suitable for subsequent analysis
TRUE

The main aim of whitening is to reduce data redundancy
TRUE


The main aim of using SIFT for feature extraction is to obtain features that are very sensitive to changes in scale, rotation, image resolution, illumination, etc.
FALSE

SVM is efficient on
Clear Margin Seperation


Each layer is composed of ___________, where the computation happens
nodes

CIFAR-10 image dataset consists of ____________ training data and _____________ testing data
50k 10k

PCA and SVD are dimensionality reduction techniques
TRUE

Cross Validation is performed on ________________________
Unknown Data

ZCA stands for ___________
Zero Phase Component Analysis

________________________ is the number of correct predictions that the occurrence is positive
False Negative x

In cross validation, the number of samples used for training the model is _____________
decreased

Data Partitioning will help us to know efficiency of model
TRUE

Neural network consistis of __________ different layers
3

Scikit Learn is an machine learning python package.
True

Allowing training data to be included in testing data will give actual performance results
FALSE

CNN is mainly used in
Image Recognition

SIFT stands for ________________/
Scale Invariant Feature Transform