What could be the task of the final project?

[metinsenturk]

The following are my suggestions.

1. Computer vision as the main area. Each group will do a task within the computer vision. For example, the following could be tasks.

Group 1 => Object Detection Group 2 => Motion Detection Group 3 => Optical character recognition (OCR)

What else do you think? Suggest and provide a detailed answer to the area you want to work on. Groups can do separate tasks as well. Reply to this issue!

[Paulkgithub123]

I am not familiar with the above, can we discuss it in the next class?

[bchavan22]

I couldn’t understand properly, request please discuss in class.

[nachammai24]

I am also not familiar with the above topics. I gathered the below details about the above topics. 1 : Optical Character Recognition for Banking Industry. Optical character recognition (OCR), is software that identifies and extracts text from documents and images, allowing them to be understood and processed by a computer. OCR has existed for a while already, but is only recently being expanded into broader data capture technology. Accurate OCR and/or data capture technology is a key component of almost any business process automation (BPA) system, which is being used across nearly every industry, to automate all types of departments and processes. The banking industry was one of the first, and remains one of the largest users of OCR and automation technologies. Data capture and automation technologies have simplified banking processes and made them immeasurably faster and more efficient. ATMs, or automated teller machines, were one of the first, and are still one of the most well-known examples of automation. They also implemented the first and most rudimentary OCR technology, used to recognize the numbers at the bottom of checks. In fact, check numbers have their own font specifically to make the numbers more easily distinguishable to a computer. The technology has since advanced to include computer vision and computer perception capabilities, which are now used to process all types of forms, applications and other paperwork that credit card processors are responsible for. Today, advanced OCR enables credit card processors to utilize state-of-the-art merchant statement analytics, which provides a true effective rate based on accurate readings of merchant transactions. OCR has also enabled customers to scan and deposit checks from their mobile phones with a machine reading and processing every part of the check. This includes items like the account number, the dollar amount and the signature. Accurately understanding scanned forms, checks and other document and images, is the first critical step in any full banking automation solution. 2 : Object Detection : Healthcare Object recognition, especially its deep learning-based variety, has the potential to save lives by helping doctors diagnose diseases from high-resolution photographs, MRIs, and CT scans. Medicine is a rare field that already possesses well-organized and labeled collections of images, making deep learning model development faster and easier. 3 : Motion detector : A motion sensor (or motion detector) is an electronic device that is designed to detect and measure movement. For example, motion sensors can be used to activate floodlights, trigger audible alarms, activate switches, and even alert the police. I am interested in Finance/Banking, Healthcare domain for the project. Lets have further discussion to understand the expectation.

[komali2208]

As per my findings the information I got about the above topics:

1. OBJECT DETECTION: Object Detection is a technology in which the computer vision deals with identifying and locating certain types of images( humans, cars, etc..). The identified images can be interpreted in various ways. For example: By creating a box around images, marking every pixel in the image which contains objects and more.

2. MOTION DETECTION: Motion Detection detects the change in position of an object relative to its surroundings or a change in the surroundings relative to an object.

3. OPTICAL CHARACTER RECOGNITION: OCR is the electronic conversion of text of images, handwritten text or printed text into machine encoded text( into digital text).

Please correct me if I am wrong. I would also like to get more detailed information of these topics so that I could choose the topic for my project.

[aakula7]

I would like to do Object Detection or Optical Character Recognition.

My project proposals for the two topics:

Object Detection -> Currently China is using object detection to monitor people with COVID-19 and who are not taking the necessary precautions such as being out in public, wearing masks, etc. Although it is an intrusion on civil liberties, many people are opting for monitoring and it forces people to take the necessary precautions.

Optical Character Recognition -> I would like to use NLP to analyze company shareholder meetings to predict the future of companies and how the company will be doing as well as giving the sentimental analysis for investors.

Please let me know if these are projects we can do

[GulkhanAnassova]

I found some information about computer vision types and I added my understanding below.

Object detection detects semantic objects such as the face, pedestrian detection, human and buildings in digital images and videos. Methods for object detection are machine learning or deep learning - based approaches.

Motion detection detects changing movements around an object such as optics(camera & video), radio frequency energy, sounds, vibrations, and magnetism. MD has two methods: mechanical(switch or trigger where we could switch manually off or on) & electronic(it is a sensor where detects optical and acoustic detection).

Optical character reader detects electronic or mechanical conversion of images, printed or handwritten texts(such as scanned or a photo of documents). OCR uses mostly for data entry(bank statements, checks, receipts, and invoices), in airports(information extractions), and assistive technology for blind and visually impaired people.

I would like to use an education or healthcare datasets for my project to predict education quality or stroke probability. I want to start first explore the datasets(how it looks like), cleaning data and training data. I want to use sentimental analysis with plots to see the differences. I have heard a lot about facial recognition to recognize people who keep coming to the same place and video should automatically show his/her preferences of products. Honestly, I am not familiar with these computer vision types such as the object, motion detection, and OCR. Overall, we need your help and guidance for these projects.

[metinsenturk]

I am not familiar with the above, can we discuss it in the next class?

hi, @Paulkgithub123 and @bchavan22, there are good examples among your colleagues. For example, @GulkhanAnassova, @komali2208 have nice contributions to this discussion board.

Basically, we are researching areas that we are interested in, and in the end, one of these topics will be our final project. So, share an area that you are interested to work on!

[metinsenturk]

PS: You can make your comments visually appealing!

You can also create headers, lists, checkboxes in your comment.

# H1 Big Heading
## H2 Heading

H1 Big Heading

H2 Heading

- List item 1
- List item 2

My list:

• List item 1
• List item 2

- [x] checkbox
- [ ] checkbox

• [x] checkbox
• [ ] checkbox unchecked

For more information on markdown writing, check out this guide.

[Paulkgithub123]

I would prefer either Group 1 or Group 2, Object Detection, Motion Detection for the final project, but I am not particularly partial to any subject.

[metinsenturk]

The final project checklist!

• [ ] Find a dataset (at least 100M rows)
• [ ] Find a machine learning task

Google Datasets