Object 365 Dataset

[ASLManasa]

ML-Crate Repository (Proposing new issue)

:red_circle: Project Title : Object 365 Dataset :red_circle: Aim : To classify images captured from the camera and detect objects present in the image. Object detection deals with recognizing which object is present in the image along with the coordinates of the object. :red_circle: Dataset : https://www.kaggle.com/c/open-images-2019-object-detection/discussion/94334 :red_circle: Approach : Try to use 3-4 algorithms to implement the models and compare all the algorithms to find out the best fitted algorithm for the model by checking the accuracy scores. Also do not forget to do a exploratory data analysis before creating any model.

Hello, ML-Crate contributors, this issue is only for the contribution purposes and allocated only to the participants of SWOC 2.0 Open Source Program.

---

📍 Follow the Guidelines to Contribute in the Project :

• You need to create a separate folder named as the Project Title.
• Inside that folder, there will be four main components.
  • Images - To store the required images.
  • Dataset - To store the dataset or, information/source about the dataset.
  • Model - To store the machine learning model you've created using the dataset.
  • requirements.txt - This file will contain the required packages/libraries to run the project in other machines.
• Inside the Model folder, the README.md file must be filled up properly, with proper visualizations and conclusions.
• A Demo Project has been created to guide you about the whole structure of presenting the Project in this repository. Here's the link of the Demo Project - https://github.com/abhisheks008/ML-Crate/tree/main/Project-Demo-Folder

---

:red_circle::yellow_circle: Points to Note :

• The issues will be assigned on a first come first serve basis, 1 Issue == 1 PR.
• "Issue Title" and "PR Title should be the same. Include issue number along with it.
• Follow Contributing Guidelines & Code of Conduct before start Contributing.
• This issue is only for 'SWOC' contributors of 'ML-Crate' project.

---

:white_check_mark: To be Mentioned while taking the issue :

• Full name :
• GitHub Profile Link :
• Participant ID :
• Approach for this Project :
• Are you a participant of SWOC 2.0?
  • [ ] YES
  • [ ] No

---

Happy Contributing 🚀

All the best. Enjoy your open source journey ahead. 😎

[Pratyush-IITBHU]

Full name: Pratyush Singh GitHub Profile Link: https://github.com/Pratyush-IITBHU Participant ID: 704 Approach for this Project: I will try Faster R-CNN, R-CNN, Yolo-v3, and Yolo-v4 darknet(using both prebuilt and own custom model). And the best-performing model will be finalized. Are you a participant of SWOC 2.0? YES

[Pratyush-IITBHU]

Sir, are you not satisfied with my approach?

I can try to learn more algorithms and model techniques if above mentioned are very basic for this project.

Thank You

[abhisheks008]

Issue assigned @Pratyush-IITBHU

[Pratyush-IITBHU]

Respected Sir,

                 I have prepared the RPN and ROI pooling class for faster Faster RCNN, but when I tried to download the dataset from the Kaggle source, it asked me for account verification via mobile number.

But no matter which mobile number I wrote for verification, it is giving the message "The number can't be verified". I even deleted and re-signed-in to my account. But still, it's showing the same error. I even messaged the support, but no reply has been sent. I am trying for 3 days to extract the data but nothing is happening.

I kindly request you to help me with the issue.

Thank You

[abhisheks008]

@Pratyush-IITBHU you can use other dataset by your choice and jaught down the project. No issues. As this problem is keep popping up whenever you want to call the API and database rejects your call by showing an error message. Many people are facing this, hence you can use different dataset for this project.

[Pratyush-IITBHU]

ok sir

[abhisheks008]

@Pratyush-IITBHU what's the update here?

[Pratyush-IITBHU]

sir,

I will send it in 4 days. Actually, the preparation of large data was very time consuming and cumbersome.

It is almost completed.

Thank you

[Pratyush-IITBHU]

Sir, I need your help.

While Training the faster R-CNN model, I got stuck in this line. I really don't know why is it happening. Because I checked, none of the variables declared here is none!

I really don't know why it is showing this error

.

[abhisheks008]

Check the line no. 67 of your utils file, there must be some None datatype used.

[Pratyush-IITBHU]

but, how I can check it??

I mean I should not change the built-in codes of our environment??

I don't really know much about the backend python codes.

[abhisheks008]

@Pratyush-IITBHU have you resolved your problem/error?

[Pratyush-IITBHU]

Sir, actually I am facing a lot of errors, I don't know this is happening, sometimes it's a version error, GPU-CPU TensorFlow problem, theono-TensorFlow(assertion error) Keras backend problem, etc. I have even posted my problems on StackOverflow and discords, but the rectification of one problem gives a newer one. Once My code even got compiled without any error but its performance was way too low.

I prepared the code 10 days before, and since then I am debugging. May this be because I don't have much knowledge of neural networks because I am just a newbie in DL. Also, the dataset is larger to run on the collab.

If you permit please give me the time of some more days, if I am unable to complete then you could assign it to any other experienced contributer. There will be no problem to me.

[Pratyush-IITBHU]

some errors are even so complicated that are not even answered on the web !! like this one:

InvalidArgumentError: OpKernel 'SparseConcat' has constraint on attr 'T' not in NodeDef '[N=0, concat_dim=0]', KernelDef: 'op: "SparseConcat" device_type: "CPU" constraint { name: "T" allowed_values { list { type: DT_UINT64 } } }' [Op:SparseConcat]

[abhisheks008]

Did you able to solve the problem? @Pratyush-IITBHU

[abhisheks008]

I'm unassigning this issue from you, as you are unable to solve the error and the issue. @Pratyush-IITBHU

[pawaspy]

Full name : Pawas Pandey GitHub Profile Link : github.com/pawaspy Participant ID : na Approach for this Project : I will use open cv for the detection of object. What is your participant role? (Mention the Open Source Program name. Eg. HRSoC, GSSoC, GSOC etc.) : IWoC

[abhisheks008]

Previous PR needs some changes to do.