Hi all, hope all of you are fine, and doing great!
We are working on an advanced Vehicle Damage Detection project, which involves Car Detection, Position Identification, different types of Damages Segmentation, and even more…
Right now, we have configured CVAT on AWS, but I wanted to know, how much it will cost us if we use Roboflow for our Annotation… We expect to have around 5 annotators, and images are to be sourced from S3, for that, I just went through a blog, which showed, how can we use Lambdas for syncing images to Roboflow…?
Also, down the line, if we use Yolov5-Instance Segmentation, will we be able to do Intelligent Labelling, so that over the time, we can decrease human annotation time?
lastly, but very important, how Roboflow can help us in evaluation of our models, something like FiftyOne Integration, etc?
Thanks, for helping me…
Best Regards
When you say “evaluation,” what metrics or insights are you looking for? How well the model performs on data it hasn’t seen? How well the model performs on your test set? Your model training metrics and training graph? We have options for this:
https://help.roboflow.com/dataset-health-check-improving-your-dataset- : our Dataset Health Check tool is great for helping you to prepare a better dataset, and improve model training metrics, and performance. At the 2:21 mark, I also go over how you can use the “Visualize” tool from models trained with Roboflow Train to visualize Ground Truth labels vs. Model Predictions (available for your test set, and valid set).
If you are working on a project for implementation at a business, or as a proof of concept for implementation at a business or for a client, then you will want to file a sales form if you haven’t already: Contact Roboflow Sales
Our team will set up a session to learn more about your needs, and get you custom pricing based on your needs. There are more questions they’ll want to cover with you, such as required implementation support, devices you are deploying to, etc. More specificity helps to ensure you get our best available pricing for your requirements in this project, and future ones.
If this is going to remain a non-commercial project, such as a personal project, or a student or research project, then you can work in a private environment if you file this form and meet the criteria of our reviewers: Contribute to the Computer Vision Community
No matter the option you choose, and the form you file, we’ll be sure you receive the proper resources.
Hi @Mohamed
Thanks fro helping me. And, I am really sorry, for such late reply. But, i don’t know, I didn’t get any notification, nor any mail, of your message…
Now, I had scheduled a meeting with RoboFlow team,
But,no one joined from Roboflow, I don’t know why.
Can you tell me, if there is any option, where I can have my whole Roboflow set with AWS s3, and I use it for Annotation, and Training, an also I host my trained model for API predictions… Right now, we cannot afford much, but we can go for around 400 USD/month…
Also, I would be very grateful, if you can share with me, any resources(code/github) which I can use to convert COCO jsons or CVAT 1.1 Instance Segmentation annotations to YOLOv5 Instance Segmentation formats… Right now, only way is that I upload to Roboflow, both images, and annotations, and then convert from there. But for sure, there will be scripts available as well, that do it…
