Pull to refresh
305.95

Algorithms *

Everything about algorithms

Show first
Rating limit
Level of difficulty

How we made landmark recognition in Cloud Mail.ru, and why

Reading time 11 min
Views 2.4K


With the advent of mobile phones with high-quality cameras, we started making more and more pictures and videos of bright and memorable moments in our lives. Many of us have photo archives that extend back over decades and comprise thousands of pictures which makes them increasingly difficult to navigate through. Just remember how long it took to find a picture of interest just a few years ago.

One of Mail.ru Cloud’s objectives is to provide the handiest means for accessing and searching your own photo and video archives. For this purpose, we at Mail.ru Computer Vision Team have created and implemented systems for smart image processing: search by object, by scene, by face, etc. Another spectacular technology is landmark recognition. Today, I am going to tell you how we made this a reality using Deep Learning.
Read more →
Total votes 45: ↑44 and ↓1 +43
Comments 0

AI-Based Photo Restoration

Reading time 7 min
Views 18K


Hi everybody! I’m a research engineer at the Mail.ru Group computer vision team. In this article, I’m going to tell a story of how we’ve created AI-based photo restoration project for old military photos. What is «photo restoration»? It consists of three steps:

  • we find all the image defects: fractures, scuffs, holes;
  • we inpaint the discovered defects, based on the pixel values around them;
  • we colorize the image.

Further, I’ll describe every step of photo restoration and tell you how we got our data, what nets we trained, what we accomplished, and what mistakes we made.
Read more →
Total votes 34: ↑33 and ↓1 +32
Comments 4

Real-time edge detection using FPGA

Reading time 8 min
Views 15K

Introduction


Our project implements a real-time edge detection system based on capturing image frames from an OV7670 camera and streaming them to a VGA monitor after applying a grayscale filter and Sobel operator. Our design is built on a Cyclone IV FPGA board which enables us to optimize the performance using the powerful features of the low-level hardware and parallel computations which is important to meet the requirements of the real-time system.


We used ZEOWAA FPGA development board which is based on Cyclone IV (EP4CE6E22C8N). Also, we used Quartus Prime Lite Edition as a development environment and Verilog HDL as a programming language. In addition, we used the built-in VGA interface to drive the VGA monitor, and GPIO (General Pins for Input and Output) to connect the external hardware with our board.


ZEOWAA FPGA development board

Read more →
Total votes 55: ↑41 and ↓14 +27
Comments 46

Authors' contribution