Skip to main content

Week 9 [10.12-16.12.18] Computer vision

Computer vision gaining more popularity and with time we could replace a human with a computer in some areas. This vision not accurate enough, it's still not comparable to the human visual system, but in the future, computer vision could be even better. Sometimes people don't notice the little things in such area as medicine, where mistake could take away a someone's life.
Watch the video above and answer the questions:

  1. Have you ever heard or maybe done your own program related to image classification?
  2. In which areas could this technology be useful?
  3. Would you like to try to write your own program which will recognize objects in real time?

Comments

In fact those technologies are implemented in self-driving cars now. Still, those are not perfect and yet to be evolved into something reliable. A machine still requires a human to oversee it, although, the day in which it will need one no more is closer that in seems, as for me.
Yep, I've done few image classification tasks. Most of them by means of neural nets, although I tried couple of old technics for image segmentations using various morphological transformations and object detection algorithm using OpenCV library.
There are dozens of such areas, autonomous car, face recognition for security and daily use, medicine etc.
Yes, it would be cool to implement a neural net which can recognize movements. In fact it is already done. But when you do it by yourself it gives you so pleasant feelings.
No, I've never done any program connected with image classification, but my friend created an application which was using face recognition to put meme faces onto them. It was really funny when he turned it on while being in crowded place :D I would like to create a program which would use objects recognition by myself but unfortunately I don't have any spare time for it at now.
I've heard about image classification in the context of cars and their systems which are detecting for example speed limit signs and are warning driver that he should slow down.
I think this technology can be used almost everywhere, but I would love to see it in medicine for instance while performing an endoscopy. Recognizing some abnormalities in the human body would be something remarkable.
Thank you for your reply, I agree with you that machine still requires a human to oversee it, but with time it will change. The human could do mistakes but machine with good software don't do mistakes. But we have different view of this topic.
Thank you for your reply, you had interest experience at this area. I agree with you, that when you do it by yourself it gives you good feelings. Good luck at your future programs ;)
Thank you for your reply, that's really cool application :). I hope you will find a time to do it. Yeap, I agree that this technology could help in medicine and a lot of other areas.
Yes, I have written a program related to image classification. It was one of the most interesting R&D projects in my company. To answer your third question as well, this was a project with real-time recognition.

I believe the second question is out-of-date. This technology has many applications already. It’s used in autonomous cars, medicine, military, and civil surveillance.
Unknown said…
Yes, I have heard about this area of machine learning, however I have never tried to to build my own solution for image classification.
As it was stated in the video, using image classification, we can build general-purpose object detection systems so number of its applications seems to be unlimited and depends only on creativity and knowledge of researchers and developers who will be using it. The most obvious are of course autonomous driving and detecting cancer cells in medical images .
Definitely, I would like to play a little bit with image processing, but I would rather start with some basic applications like classifying static images because detecting objects in real time seems to be too advanced topic to start with.
Technology could help in almost every area IMO. I think that it's quite cool that AI is such an easy thing to do, you have a ton of programming libraries to help you. On the other hand there a ton of cancerous companies/ project that are using buzz words and do not think how exactly their project will profit from using this or that.
Thank you for your reply, it's cool that you had experience at this area. And I think this experience will be very useful in your future work. Maybe, you are right, this technology already used in a lot of areas.
Thank you for your reply, I agree with you that it depends only on creativity and knowledge. It's good idea to start from basic things, then it will be easier to understand more complicated things.
Unknown said…
No, I did not make my own, but I really like that this technology exists.
As mentioned in the video, I like the idea of ​using such a program in medicine, for example, to make the examination of internal organs less painful. Even in the very first minutes of the video, it seemed to me that such object recognition could help in advertising; by the principle of recognition, one can immediately understand to whom what to offer.
This is an interesting topic, I will think about it because I have never had such an idea.

Thank you for your reply, I think this technology has a big future. If you have some time and this topic interesting for you, you definitely need to try and to my mind it will be very useful for you.
My engineering thesis is actually about Object Classifications in computer vision. I'm writing my own program to classify objects and name them. This is probably the easiest kind of computer vision problems but also the most basic. Once you can classify objects, you can start looking at a bigger picture and find them there.
This technology is very exciting field for me - right now there is a lot of research going on. People are looking for faster ways to detect objects, to classify them and later to take actions based on this classification. There are a lot of such products already used in various ways: face detection on pictures, fingerprint classification, security (CCTV analysis), medicine (cancer detection) and many many others. I hope I will be able to build new products using computer vision in near future.
I never wrote my programs. It seems to me that this is incredibly difficult and more look like magic. But the recognition of objects in the image is incredibly cool. Such software will be used everywhere. Starting from security systems to various games and simply using civilians for their own purposes.
I had some minor experience with image classification during one of the earlier classes, where we tried to build robots that would follow red balls dangled in front of them. This technology is progressing in an incredibly fast rate, and seeing the video in the post gives me hope that it will soon be implemented in real world applications. The mention of possible use in medicine is great news, as much as any sort of monitoring that could lead to more safety for people. An example brought up in one of the earlier posts was using drones with appropriate monitoring software to spot sharks swimming close to beaches in Australia. Playing around with image recognition is something that’s certainly on my “to do” list, but I’m focusing on other things at the moment so don’t have much time to touch that.
Have you ever heard or maybe done your own program related to image classification?
Yes I heard about it before but I don't have any personal experience in dealing with this technology.
In which areas could this technology be useful?
As mentioned in the video: healthcare, self-driving cars. However, I think that this technology is already used in smartphones for face recognition. I think it could be used in such areas as military for recognizing targets or security. For example in detecting a burglar.
Would you like to try to write your own program which will recognize objects in real time?
I think that this is quite a interesting topic overall. However, I have never been really focused on it. I am more interested in other things and I would prefer to work on them.
Thank you for your reply, I agree that better to start from basic things. Then will be easier to do more complicated programs. This technology will be very useful for you and I think that in near future will be a lot of job offers based on this technology.
Thank you for your reply, programming isn't magic, you need to spend a lot of time to learn how to write programs. But it's cool feelings when you have done your own program.
Anna Moskalenko said…
I’ve never done something like that, it sounds very complicated to me, I don’t think that I’m that smart :D But my friend doing this kind of project for his diploma, so yeah, I’ve heard.
Well, in all areas, I think. Starting from the medicine or even this funny project, that I’ve heard about some time ago, a program which can show you brands of clothing that for example a person in you favorite TV series wears and where you can buy it. Great thing!
Well, I kinda have it done already, but this program is very simple and can recognize only one object that I told it to recognize :) Again, it’s very complicated and I think it’s not for me.
Thank you for your reply, you had interesting experience. In IT a lot of interesting things. But maybe with time you will change your preference.
Yes, I have heard about computer vision and done a few programs in this area. For me DL(Deep Learning) and especially image classification is one of the favorite topics in AI.
This technology is already useful in a lot of areas such as medicine, self-driving cars, fingerprint detection and others.
I have already wrote few programs in image classification area and also working on two more ones. One of them is connected to my graduate work: real time detection and recognition. And the other one is connected to my work, that’s why I can’t tell you more.
Aleksander Wiza said…
I was working on this project that decided wether a picture was a cat, a dog or undecidable. However, it was a part of an online course dealing with classification problems in machine learning.

2. One day it will be used in every industry possible, the methods behind image recognition can be used in differently structured problems.

3. I'd love too.
Thank you for your reply, in what things are you interested? In IT a lot of different interesting tasks. But to my mind at this moment the Computer vision has become a very popular topic.
Unknown said…
Yes I heard and even wrote a basic program which follow the previously given color. It was very interesting experience to create such thing.
We can create for example program for blind people to recognize things and telling them what is in front of them. Second thought is creating futuristic robot which can help disabled people in home and outside.
Szymon Gibalski said…
1. Have you ever heard or maybe done your own program related to image classification?
- No, I have not done any program related to image classification, but I would really like to do something in this area.
2. In which areas could this technology be useful?
- For example, an application which recognized cars.
3. Would you like to try to write your own program which will recognize objects in real time?
- Yes, like I mentioned before I would like to do something with this technology. Because it is really interesting and would look really nice in my CV.
Marcin Zając said…
1. Have you ever heard or maybe done your own program related to image classification?
Yes, I'm writing an app, which use object detetction.

2. In which areas could this technology be useful?
This technology can be used to help people with low vision or completally blind.

3. Would you like to try to write your own program which will recognize objects in real time?
I'm working on application, which recognize buses and their numbers in real time.
Unknown said…
I'm working on it in my job. We're trying to recognize the situation on image/photography by recognizing objects and environment. We're creating photobooks and by recognizing the objects, we're trying to suggest which template should be used etc.

Yes, I'd like to and maybe someday I will try to ;).

Popular posts from this blog

Week 1 (09-15.03) VOD

http://www.vod-consulting.net/wp-content/uploads/2012/08/1.jpg

Week 11 [03-09.06.2019] The problem with ecological cars emission in UK

The problem with ecological cars emission in UK Since the adoption of the European Emission Allowance Directive in the European Parliament, all car makers have tried to submit. Since 1992, the Euro I standard has been in force, which limited the emission of carbon monoxide to the atmosphere. The Euro VI standard currently applies, which limits the series of exhaust gases. These include: hydrocarbons, nitrogen and carbon oxides, and dust.   The most significant change was brought by the Euro IV standard. For the first time it introduced the limitation of nitrogen oxides, which are responsible for the harmful compounds of smog.   What is smog?   Smog consists of sulfur oxides, nitrogen and carbon. In addition, solid substances such as suspended dust (PM). Dust suspend in atmospheric aerosols may be in liquid and solid form. These can be particles of sea salt, clouds from the Sahara and artificial compounds made by people. These compounds often come fr

Week 4 [06-12.11.2017] This is what happens when you reply to spam email.

James Veitch is a British comedian. In today’s Ted Talk James with characteristic for himself a sense of humor shows how he deals with spam emails and why responding to junk messages may be sometimes dangerous. Questions: What do you think about James’s  way of dealing with spam? Why are junk messages legal, even though it sometimes may be a fraud? Dou you have a problem with spam? How do you deal with with it?