ROADAR Personal
Co-author and ideological inspirer of the RoadAR project Marat Bashlykov relied on his own need to notice signs on the road. Traveling a lot by car, he noticed that the driver does not always see signs, and this threatens with fines and generally increases the danger of driving. On the other hand, many have a DVR, the camera of which is capable of fixing signs, and a phone that can process data from the camera. So the idea was to create RoadAR. Last year, the project participated in the GenerationS acceleration program, reached the final of the Web Ready contest, and this year became the winner at the Startup Village conference in the Intel nomination and is going to represent Russia at the Intel Global Challenge contest in California.
SK: Marat, tell us how RoadAR works?
MB: RoadAR (note: a phone equipped with the RoadAR application goo.gl/yC6F9t) is put on the holder in the car, starts up and starts recording video, simultaneously analyzing the image with computer vision algorithms. On a modern smartphone, with a frequency of 30 frames per second, it detects and recognizes road signs, understands their zone of action and warns the driver in case of a violation or of a possible danger. There are a lot of difficulties here: getting a smartphone to simultaneously record video and analyze it is no longer easy, but we also need to impose coordinates on the video (i.e. mix it), perform recognition, so that it works on a huge fleet of android devices. The biggest problem now is power consumption, but soon we expect to significantly reduce it by optimizing the performance of recognition algorithms.
')
SK: Do you use any third-party development?
MB: We use OpenCV for some basic transformations and operations, but we write a lot ourselves to make it work fast enough on mobile devices. A lot of work has been done in the field of computing on the GPU, but so far we are not connecting it because of the same power consumption.
SK: How accurately does the system work? What affects the quality of recognition?
MB: There are 2 key points in the recognition of characters (if we assume that the smartphone is quite powerful and it has a good camera) - the detection of a candidate for a sign and, in fact, recognition itself. Naturally, in the case of a smartphone, we cannot run through the window throughout the frame and look for signs there. Now we are looking for suitable forms and, when they are found, we give them to “recognizer”. We have achieved a quick search for basic forms, but the only limitation now is that they should be closed. Those. the overlap often makes it difficult to find a form, but the shadow does not. On the GPU, it turns out to look for any form and quickly enough, but the power consumption grows even stronger. Now we are working closely on a fast algorithm for finding forms on the CPU, which overlap would not have so much affected.
The accuracy of the recognition of candidates for most characters is very high, for some it is 99.7%, for some 95%, but, of course, there are difficulties. For example, the sign "give way" - he is often confused with the wires and masts of power lines. The fact is that this is the only sign that is an inverted triangle and it has no special signs by which it could be distinguished and for which the algorithm would get hooked. We have a lot of ideas on how to solve this problem, now we are working on their testing.
SK: How is the monetization model built?
MB: We make all the basic functionality free. Now we are thinking about creating several additional chips for money or for a share, for example - a warning about traffic lights and holes. We can also collect a lot of information that is interesting to different businesses. Information from billboards, prices at gas stations, etc. We also collect specialized signs for trucks, such as signs such as "weight limit", "height restriction", "weight limit per vehicle axle", etc., "the movement of trucks is prohibited." All this information is necessary for drawing up the optimal route and logistics companies have already been interested in us with similar data. You can collect a large amount of various information - about traffic jams, accidents, the quality of road infrastructure.
Another idea is advertising. Sound advertising, which should be interesting to the driver. For example, in the morning and in the evening you stop at a kindergarten, therefore, you have a child and you may be interested in advertising of children's goods, respectively, when approaching the children's world you will hear an offer from a store, and even containing a discount. Advertising can be useful and it should not be much.
SK: Tell us a little about the project team?
MB: We have a small team, 5 people. Yevgeny and I travel a lot by cars, by previous work we often wandered on business trips, and now we often go to competitions in other cities, and always by cars. Faced with the fact that modern gadgets do not fully use their capabilities. Constantly thought and discussed how to improve them. We decided that step by step we will be able to do a lot ourselves, and we will be able to collect data for cartography using crowdsourcing. They quit their jobs, put together a team and sat down to work. Yevgeny called his former classmate Andrei to us, who at that time already had a successful business, he worked on outsourcing with the USA and Australia, making high-load decisions. He was also interested in our idea, and so much so that he left his business to partners and came to our team.
Two more very cool guys - Foat and Sasha. Foat is an expert in computer vision and machine learning, we were looking for him for a long time. We went to the universities talking with professors, they advised us to their students, but someone was afraid to leave their homeland, someone did not suit us for other reasons. And when we met Foata, he was imbued with the idea, and he was not the driver himself, he was just curious to do what we are doing.
And we met Sasha by pure coincidence, so to speak, a coincidence: in Kazan IT Park there was a job fair; we didn’t really hope for her, but closer to completion a young guy came up to us, said he didn’t know what could be useful for us, but he was interested in drank such a thing. So it works with us now, without him we are as without hands :)
At the very beginning, we were all told that what we had in mind was impossible to do, smartphones would not cope with recognition. For us, this has become a kind of challenge, and now we periodically prove that nothing is impossible, especially when such a harmonious team does it!
SK: How did you manage to find investments?
MB: We have big problems with investments. We do not online store where everything is clear: sales funnels, landings, CPA, LTV.
We have a complex monetization model and from the technological side everything is very sophisticated. Apparently, many investors are at risk. For half a year we did everything with our own money, invested a fairly large amount of personal funds. Then, three angels and the MSU Business Incubator invested in us micro investments. This money was enough to release the official release and understand that people are interested in what we do. Last week we received 15K installations. Now we are looking for the next round of investment.
SK: Tell us about your impressions of participation in GenerationS?
MB: The most wonderful impressions are a lot of good projects, interesting speakers, useful lectures. As a result, this is an excellent networking and the opportunity to talk with representatives of large companies such as Intel, Yandex, listen to David Jan and Natalia Kasperskaya.
SK: “Where to run” to a young startup manager who has an idea, but no money / qualifications to implement it on his own?
MB: Into the wall. Wherever you go, get ready that there will be a wall in front of you. But, if you believe in your idea - do not give up :)
We started with the fact that we came to the Kazan IT Park Business Incubator, and this helped us a lot. Now there are many interesting acceleration programs, incubators, etc. There you can check your idea, find like-minded people, and even, perhaps, attract small investments for launching.
Illustrations:
Video:
Interview for the GenerationS contest was taken by Sergey Kokarev
SK: Marat, tell us how RoadAR works?
MB: RoadAR (note: a phone equipped with the RoadAR application goo.gl/yC6F9t) is put on the holder in the car, starts up and starts recording video, simultaneously analyzing the image with computer vision algorithms. On a modern smartphone, with a frequency of 30 frames per second, it detects and recognizes road signs, understands their zone of action and warns the driver in case of a violation or of a possible danger. There are a lot of difficulties here: getting a smartphone to simultaneously record video and analyze it is no longer easy, but we also need to impose coordinates on the video (i.e. mix it), perform recognition, so that it works on a huge fleet of android devices. The biggest problem now is power consumption, but soon we expect to significantly reduce it by optimizing the performance of recognition algorithms.
')
SK: Do you use any third-party development?
MB: We use OpenCV for some basic transformations and operations, but we write a lot ourselves to make it work fast enough on mobile devices. A lot of work has been done in the field of computing on the GPU, but so far we are not connecting it because of the same power consumption.
SK: How accurately does the system work? What affects the quality of recognition?
MB: There are 2 key points in the recognition of characters (if we assume that the smartphone is quite powerful and it has a good camera) - the detection of a candidate for a sign and, in fact, recognition itself. Naturally, in the case of a smartphone, we cannot run through the window throughout the frame and look for signs there. Now we are looking for suitable forms and, when they are found, we give them to “recognizer”. We have achieved a quick search for basic forms, but the only limitation now is that they should be closed. Those. the overlap often makes it difficult to find a form, but the shadow does not. On the GPU, it turns out to look for any form and quickly enough, but the power consumption grows even stronger. Now we are working closely on a fast algorithm for finding forms on the CPU, which overlap would not have so much affected.
The accuracy of the recognition of candidates for most characters is very high, for some it is 99.7%, for some 95%, but, of course, there are difficulties. For example, the sign "give way" - he is often confused with the wires and masts of power lines. The fact is that this is the only sign that is an inverted triangle and it has no special signs by which it could be distinguished and for which the algorithm would get hooked. We have a lot of ideas on how to solve this problem, now we are working on their testing.
SK: How is the monetization model built?
MB: We make all the basic functionality free. Now we are thinking about creating several additional chips for money or for a share, for example - a warning about traffic lights and holes. We can also collect a lot of information that is interesting to different businesses. Information from billboards, prices at gas stations, etc. We also collect specialized signs for trucks, such as signs such as "weight limit", "height restriction", "weight limit per vehicle axle", etc., "the movement of trucks is prohibited." All this information is necessary for drawing up the optimal route and logistics companies have already been interested in us with similar data. You can collect a large amount of various information - about traffic jams, accidents, the quality of road infrastructure.
Another idea is advertising. Sound advertising, which should be interesting to the driver. For example, in the morning and in the evening you stop at a kindergarten, therefore, you have a child and you may be interested in advertising of children's goods, respectively, when approaching the children's world you will hear an offer from a store, and even containing a discount. Advertising can be useful and it should not be much.
SK: Tell us a little about the project team?
MB: We have a small team, 5 people. Yevgeny and I travel a lot by cars, by previous work we often wandered on business trips, and now we often go to competitions in other cities, and always by cars. Faced with the fact that modern gadgets do not fully use their capabilities. Constantly thought and discussed how to improve them. We decided that step by step we will be able to do a lot ourselves, and we will be able to collect data for cartography using crowdsourcing. They quit their jobs, put together a team and sat down to work. Yevgeny called his former classmate Andrei to us, who at that time already had a successful business, he worked on outsourcing with the USA and Australia, making high-load decisions. He was also interested in our idea, and so much so that he left his business to partners and came to our team.
Two more very cool guys - Foat and Sasha. Foat is an expert in computer vision and machine learning, we were looking for him for a long time. We went to the universities talking with professors, they advised us to their students, but someone was afraid to leave their homeland, someone did not suit us for other reasons. And when we met Foata, he was imbued with the idea, and he was not the driver himself, he was just curious to do what we are doing.
And we met Sasha by pure coincidence, so to speak, a coincidence: in Kazan IT Park there was a job fair; we didn’t really hope for her, but closer to completion a young guy came up to us, said he didn’t know what could be useful for us, but he was interested in drank such a thing. So it works with us now, without him we are as without hands :)
At the very beginning, we were all told that what we had in mind was impossible to do, smartphones would not cope with recognition. For us, this has become a kind of challenge, and now we periodically prove that nothing is impossible, especially when such a harmonious team does it!
SK: How did you manage to find investments?
MB: We have big problems with investments. We do not online store where everything is clear: sales funnels, landings, CPA, LTV.
We have a complex monetization model and from the technological side everything is very sophisticated. Apparently, many investors are at risk. For half a year we did everything with our own money, invested a fairly large amount of personal funds. Then, three angels and the MSU Business Incubator invested in us micro investments. This money was enough to release the official release and understand that people are interested in what we do. Last week we received 15K installations. Now we are looking for the next round of investment.
SK: Tell us about your impressions of participation in GenerationS?
MB: The most wonderful impressions are a lot of good projects, interesting speakers, useful lectures. As a result, this is an excellent networking and the opportunity to talk with representatives of large companies such as Intel, Yandex, listen to David Jan and Natalia Kasperskaya.
SK: “Where to run” to a young startup manager who has an idea, but no money / qualifications to implement it on his own?
MB: Into the wall. Wherever you go, get ready that there will be a wall in front of you. But, if you believe in your idea - do not give up :)
We started with the fact that we came to the Kazan IT Park Business Incubator, and this helped us a lot. Now there are many interesting acceleration programs, incubators, etc. There you can check your idea, find like-minded people, and even, perhaps, attract small investments for launching.
Illustrations:
Video:
Interview for the GenerationS contest was taken by Sergey Kokarev
Source: https://habr.com/ru/post/225529/
All Articles