search

OlafenwaMoses / ImageAI

A python library built to empower developers to build applications and systems with self-contained Computer Vision capabilities
https://www.genxr.co/#products
MIT License
8.63k stars 2.19k forks source link

Custom Training for Object Detection #89

Closed JimZuber closed 6 years ago

JimZuber commented 6 years ago

I really like what you have done with ImageAI, it has been very helpful getting me started doing hands on work with AI. Question...

It appears that the latest release supports custom training for Image detection, but not object detection. Is that on your roadmap for imageAI in the near future? I am unclear whether there are additional complexities to training the model for object detection.

Jim

JimZuber commented 6 years ago

I see that you have commented on this topic in your responses to another issue, so sorry about the duplicate post. Nevertheless, it would be great if you could provide an update/estimate for this feature.

OlafenwaMoses commented 6 years ago

Yes, We have plans for adding support for training custom detection models. This we are planning to announce along with a new tool we are building called "DeepStack" . DeepStack is an easy to setup, self-hosted, offline and programming language independent Artificial Intelligence server. DeepStack can be described as Google Vision Cloud sitting on your Laptop and can be queried by any device around it.

We are running a DeepStack developer survey to build the right features into it and by filling the survey, you get signed up for the DeepStack Insider Preview program taking place in December where you get to try the first version of DeepStack completely free with one-on-one support, tutorials and documentation provided. Visit the link below if you want to have first-hand experience of DeepStack.

https://docs.google.com/forms/d/e/1FAIpQLSd7SLc99snRE53VNVkDkRBTwu2MH2_5vCCC6YqAxkIlAUzlrw/viewform

OlafenwaMoses commented 6 years ago

For further discussion on this, please refer to #8