98 points by sephiroth74 7 months ago flag hide 12 comments
user1 7 months ago next
Nice work! I've been looking for a real-time image segmentation API for my new project. How accurate is the segmentation? And what kind of input formats do you support?
creator 7 months ago next
Thanks! The segmentation is pretty accurate with an average IoU score of 0.85 for most images. Currently, we support JPEG and PNG formats, but we're planning to add more soon.
user2 7 months ago prev next
0.85 IoU is impressive, well done! Also, what kind of hardware and libraries do you use to run this service? I'm thinking of creating something similar and could use some tips.
creator 7 months ago next
We use Nvidia RTX 3090 GPUs and our backend is powered by FastAPI and PyTorch. I'll be writing a blog post soon about the infrastructure and implementation details, so stay tuned!
user3 7 months ago prev next
How much do you charge for the API usage? And do you offer any free tiers or credits for new users? I'd love to test it for my research project.
creator 7 months ago next
We offer a generous free tier with 500 requests per day and 2,000 images per month. The paid plans start at $29/month with no throttling or image limits. You can also upgrade to the annual plan and get a 20% discount.
user5 7 months ago prev next
Is it possible to get a bulk or batch processing feature? I have a lot of images to process and handling them one-by-one might not be the best approach for me. Thanks!
creator 7 months ago next
Absolutely! We offer a batch processing feature that allows you to upload hundreds or thousands of images in one request. Just let us know if you need any guidance on how to use it and we'll be happy to help.
user4 7 months ago prev next
That's awesome, thank you! I'll definitely consider upgrading once I have a better idea of the volume of requests I need. One more question - do you offer any tools or libraries for visualizing the segmentation results?
creator 7 months ago next
Yes, we do! We offer pre-built widgets and integrations for popular data visualization libraries such as Plotly, Matplotlib, and Bokeh. You can also download the raw data and create your own custom visualizations.
user6 7 months ago prev next
Do you have any examples or tutorials on how to use the API with different programming languages? I'm a Python user myself, but I have colleagues who prefer other languages.
creator 7 months ago next
Yes, we do! We have examples, tutorials, and documentation in multiple languages including Python, Java, Go, Node.js, Ruby, and C#. Just check out our website for more details and resources.