### Weighted Loss Functions

Weighted Loss Functions For image classification and localization we need to use 2 loss functions on the same network!, to calculate the predicted class we would use categorical cross entropy but to find the bounding box (which is a regression problem) we need to use something like an SME, L1 Loss or smooth L1 Loss.…

### Visualizing Feature Maps

Visualizing Feature Maps When a Convolutional Neural Network trains, it is actually adjusting the weights of each node, then each node will become a kernel (or filter) that will be applied to an image, each filter will play a purpose, it can be to detect horizontal o vertical edges, blur out noise, detect patterns,…

### Region Proposal Algorithms

Region Proposal Algorithms R-CNN is a Region Convolutional Neural Network It is used to produce a limited set of cropped regions to analyze ROIs (Regions of Interest) The R-CNN is the least sophisticated region-based architecture The next one is the Fast R-CNN algorithm. Fast R-CNN is about 10 times as fast to train as an…

### GPU

The GPU performance state APIs are used to get and set various performance levels on a per-GPU basis. P-States are GPU active/executing performance capability and power consumption states. P-States range from P0 to P15, with P0 being the highest performance/power state, and P15 being the lowest performance/power state. Each P-State maps to a performance level.…

### Batch Size

Batch size refers to the number of training examples utilized in one step or iteration. One step or iteration is one step of gradiend decent (one update of weights and parameters) The batch size can be either: The same number of the total number of samples which makes one step = an epoch, this is called batch…

### Localization

Robot localization in essence is based in two main steps: Sense and Move It will start with a initial belief (or prior) of maximum confusion where the probability distribution will be uniform (flat, which means it has the same value everywhere) Then it will start cycling through sensor measurements (Sense) and movements (Move) When the robot moves…

### My Top 5 Recommended Places to Learn about Deep Learning and Machine Learning

Continuing on my #100DaysOfMLCode, these are some of the courses I’m following and recommend if you are interested in learning ML and DL. One of the most stunning statistics in the area of Machine Learning (ML) was released by Tractica. According to the company, ML will grow from its \$1.4 billon value as at 2016 to…

### Probability Basic Concepts

Independent Events The probability of an event does not affect the probability of the next event. So for example tossing a coin does not affect the probability of the next flip. Dependent Events Two events are dependent when the probability of one influences the likelihood of the other event. Joint Probability Is the probability…

### Motion in Computer Vision

Motion in Computer Vision Motion can be tracked with a 2D motion vector. A vector has a direction and magnitude which will determine the direction and amount of movement between one frame and the next one. First we will have to defive special points to track like for example intersections or corners once we localize…

### Recurrent Neural Networks (RNNs)

Will mantain states, the output depdends on the current input as well as previous ones. This means that the input of the next step will include their own inputs + the output of the hidden layer of the previous step (the memory cell) and we will refer to it instead of h (for hidden layer) to s for…