Tag: deep learning models
-
Intelligent Defect Detection in Metal Parts using Optical System for Industry 4.0 Manufacturing
Detecting and classifying aesthetic defects in different sizes, shapes, and positions immediately after the casting process is an essential task for the quality control unit. In this paper, we introduce a simple, low-cost, and efficient optical system powered by deep learning models to quickly, accurately, and automatically identify and classify casting defects.
-
Model-Driven Vs Data-Driven Approach
An AI model’s performance can be increased by either improving the dataset or the model’s structure. The purpose of this article is to examine these two approaches and determine which one is the most efficient.
-
TensorRT
TensorRT is a library developed by NVIDIA for faster inference on NVIDIA graphics processing units (GPUs). It can improve inference time for many real-time services and embedded applications, 4-5 folds. The optimization TRT applies to deep learning models will be examined in this article.