Rinat S, PhDinProduct AIHow to collect dataData can be obtained in three ways: downloading a ready-made dataset, getting it using the provided API, and collecting it using modern…Aug 24, 2021Aug 24, 2021
Rinat S, PhDinProduct AIAI Trends 2021–2025One of the trends of 2021 that will continue at least for the next few years is the rise in popularity of the PyTorch framework. This can…Aug 10, 2021Aug 10, 2021
Rinat S, PhDinProduct AIEnd-to-end CNN detection and descriptionLet’s explore using convolutional networks to solve problems stemming from detecting image points.Aug 5, 2021Aug 5, 2021
Rinat S, PhDinProduct AIImage data augmentationMore advanced augmentation techniques use neural and convolutional networks to generate new data. Therefore, the SMOTE approach is based on…Jul 27, 2021Jul 27, 2021
Rinat S, PhDinProduct AIPopular Scraping LibrariesOne of the ways of obtaining data is “scraping”, which can be done using modern engines such as Selenium, JSoup, and others. A notable…Jul 8, 2021Jul 8, 2021
Rinat S, PhDinProduct AIHandcrafted local featuresVarious detectors are used to search for special points in an image. Point descriptors allow you to distinguish one particular point from…Jun 10, 2021Jun 10, 2021
Rinat S, PhDinProduct AICNN Neural Networks ScalingMethods for scaling neural networks, increasing depth, resolution, and width.Jun 8, 2021Jun 8, 2021
Rinat S, PhDinProduct AICPU VS GPUAnalogs between Python packages and modules for CPU and GPUJun 3, 2021Jun 3, 2021
Rinat S, PhDinProduct AITraditional GPU computing vs RAPIDS modelUsing RAPIDS allows you to speed up the processing of data tenfold.Jun 1, 2021Jun 1, 2021