Dive into the second part of an introduction to neural networks in this comprehensive lecture from the University of Central Florida's Computer Vision course. Explore computational implementations of neural activation functions, binary image classification, and multi-class neural networks. Learn about bias convenience, composition, and the introduction of non-linearities in neural networks. Examine various activation functions, including the UCF Perceptron model, and understand the structure and goals of multi-layer perceptrons (MLPs). Conclude by evaluating MLP performance on the MNIST dataset, gaining valuable insights into deep learning applications for computer vision tasks such as filtering, classification, detection, and segmentation.
Overview
Syllabus
Administrative
Computational Implementation of the Neural Activation Function
Binary classifying an image
Neural Networks - multiclass
Bias convenience
Composition
Problem 1 with all linear functions
Let's introduce non-linearities
Activation Functions
UCF Perceptron model
Multi-layer perceptron (MLP)
Goals
MLP performance on MNIST
Taught by
UCF CRCV
Tags
Related Courses
-
Introduction to Neural Networks - Lecture 5
-
Convolutional Neural Networks: Deep Learning
-
Build Decision Trees, SVMs, and Artificial Neural Networks
-
Computer Vision: Classification Techniques - Neural Networks and CNN Architecture - Lecture 12
-
Deep Learning with Keras
-
Introduction to Neural Networks for Computer Vision - Part I
