innovator

Add idea


Calendar

«    July 2019    »
MonTueWedThuFriSatSun
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
 

 

Advert

 

Payment

 

Advert

 

Authorization

Стартап

DateDate: 29-06-2019, 06:26

In addition to embedding the camera under the display, the company also upgraded the camera module along with its suppliers, increasing the aperture, sensor and pixel size.
Introduced the first smartphone with a sub-screen camera
 
At the Mobile World Congress in Shanghai, the Chinese company Oppo presented a smartphone with a sub-screen selfie camera.
You will not surprise anyone with front-facing cameras with high resolution and two cameras on the back cover. Therefore, the Chinese company Oppo went ahead and integrated the camera directly under the touchscreen display, reports Naked Science.
 
The technology is called USC (Under Screen Camera). According to the company, it provides the effect of "absolute frameless."
The very form of the camera and its position are no different from those of a regular smartphone. It is located under the touch screen and can be seen.
This technology also allows you to simultaneously display information on the display and use the camera functions: take pictures, unlock the screen on the image of the person, and make video calls.
In addition to embedding the camera under the display, Oppo also upgraded the camera module along with its suppliers, increasing the aperture, sensor and pixel size.
The company also improved the image processing algorithm, removing the haze effect and adjusting white balance.
According to the developers, USC uses zoning technology to manage the pixels on the screen, using highly transparent material through which the light passes unhindered.
 
Oppo also demonstrated MeshTalk technology at the exhibition, which allows you to send text and voice messages, as well as make calls between Oppo devices up to three kilometers away without using cellular networks, Wi-Fi or Bluetooth.
MeshTalk unites users into local networks, which allows you to increase the range of the signal between phones without using a cellular network.

DateDate: 28-06-2019, 06:12

The new device, designed at Harvard, has become the lightest robone-insect.
Specialists at Harvard Microorobotics Laboratory have developed a robot that flies thanks to four wings and solar panels. Reported by the IEEE Spectrum.
According to the creators, robotic insects are not uncommon these days. However, most of them require strong power supplies for flying. The new device also feeds onboard source.
Robot RoboBee X-Wing has four wings, which are arranged in pairs at right angles to each other. Their scope reaches 3.5 centimeters at a height of the robot at five. The mass of the device was 259 milligrams. Of these, only 90 relate directly to the “bare” device without a battery.
Now the robot is able to fly only in a circle. For his work, the required illumination is three times greater than that provided by the sun. In the future, there will be an improved version with more convenient power consumption.

DateDate: 27-06-2019, 07:51

While smartphone makers and operators around the world are discussing the benefits of 5G, OPPO is asking another question - what if cellular communication goes down, and Wi-Fi and even Bluetooth turn off?
Answering this question, the Chinese company introduced a new technology, called MeshTalk, which can transmit data between devices up to 3 km from each other. , but also introduced MeshTalk technology. MeshTalk uses a special high sensitivity microchip to achieve long distance communication.
This is a low bandwidth system, but it supports voice calls, not just text messages. The MeshTalk technology works at a distance of up to 3 km. A local network will be created between devices that support this technology, through which you can send text messages and make voice calls up to a distance of 3 kilometers. And it is in open space.
In crowded places (at a concert or on Red Square on May 9, for example) with a large number of supported devices, the distance between interlocutors can be even greater. The new technology also consumes very little power, and therefore OPPO is already developing smartphones that can work up to 72 hours in standby mode with active MeshTalk. This technology can be useful in emergency situations when base stations are overloaded in places in places with large crowds of people or there is simply no access to the mobile network. OPPO does not disclose which technology is behind MeshTalk or when the first smartphones with it will be available on the market. A source

DateDate: 26-06-2019, 07:36

The application uses the YOLO (You Only Look Once) object classifier and a tool that can be called a context-dependent fill: it analyzes the background with the help of a neural network and fills in people’s silhouettes based on the information it received about it.
NAKED-SCIENCE 06/26/2019 00:00:11
Created an application that automatically cuts people off a photo.
 
The artist under the nickname damjanski with friends developed the application Bye Bye Camera, which automatically removes all people from the photo.
He called the program his new brush. Reported Naked Science.
“I created this project with my two colleagues, Andrey and Pavel from Russia. A couple of years ago, I gathered a group called Do Something Good, in which I connected all the people with whom I collaborated, online. Now we are 16 people, we work in different areas, but we unite to implement projects. This application removes the vanity of any selfie, and with it the person.
I see the Bye Bye Camera as a posthuman application. This is a subtle hint of the future, in which complex programs will replace people in some professions, and maybe, as some claim, will completely crowd out people. It is interesting to know what a person looks like from the point of view of Ai (yes, a little “i”).
In this case, a set of pixels that identify a person based on predefined data. But who determines this data that marks a person intangible? So many questions for such a small application, ”said the artist.
The application uses the YOLO (You Only Look Once) object classifier and a tool that can be called a context-dependent fill: it analyzes the background with the help of a neural network and fills in people’s silhouettes based on the information it received about it.
According to the creator, this application is an artistic tool rather than a business.
Its main feature is that it removes all people from a photo.
Damjanski says that a lot of friends asked him to add a function that would allow him to choose who to cut from the photo, but he continues to assert that it was in uncompromising sense.

DateDate: 25-06-2019, 07:26
It is reported by Sotsportal with reference to Hi-Tech +.
“The report, edited by David Rolnik from the University of Pennsylvania, is divided into 13 areas, from energy to agriculture and climatology,” the author noted.
The recommendations are divided into three categories: “highly efficient” for problems well suited for machine learning; “Long-term” - for solutions that do not pay off before 2040; and "risky" - methods with less predictable results. MIT Technology Review has selected recommendations from the category of the most effective.
Evaluation of the need for electricity.
If we are going to switch to renewable energy, we should more accurately calculate the need for electricity today and in the long term. Already there are algorithms that predict energy consumption, but they can be improved.
New materials.
Scientists need materials that store, collect and use energy more efficiently, but the discovery process is usually too slow and inaccurate. Machine learning could develop, for example, solar fuel that absorbs the energy of sunlight, or find a replacement for steel, which is responsible for 10% of all greenhouse gas emissions.
Optimization of freight and deliveries.
Transportation of goods around the world is a complex and often inefficient process, requiring coordination of a lot of parameters. Machine learning is able to find ways to minimize the number of journeys and associated emissions.
Acceleration of the transition to electric transport. Thanks to the algorithms, automakers can increase the range of trips on a single charge and prepare the grid for the load.
Improving the efficiency of buildings.
Smart control systems could significantly reduce the energy consumption of buildings, taking into account the weather forecast, the number of inhabitants and combining them with other factors: heating, cooling, ventilation and lighting.
Other recommendations from the same category: creation of opportunities for the development of precise agricultural technology, tracking the process of deforestation and changing the consumer's attitude to the environment to a more careful one.