Python for Computer Vision with OpenCV and Deep Learning

courses on OpenCV (an open computer vision library with easy-to-build face/object detection algos) will also be a great start

for example this course on Udemy is 11 hours and covers Objects Tracking and Face Detection: https://www.udemy.com/course/opencv-python-for-beginners-hands-on-computer-vision/

​

And this is an intermediate class on it: https://www.udemy.com/course/python-for-computer-vision-with-opencv-and-deep-learning/

Having worked with the common ML libraries on a PhD level set of applications is a great starting point. Consider going for Soledad Galli's industry-oriented ML courses on Udemy: https://www.udemy.com/user/soledad-galli/

In particular, check out her courses on Model Deployment and Testing/Monitoring ML applications, as these are highly valuable industry skills. Udemy courses are discounted regularly but you can also search online for discount codes so they are usually around $10-20 instead of $200.

If you've "only" worked with Python in the context of Jupyter Notebooks, I'd also consider learning proper coding principles, most importantly being fluent in writing in an object-oriented style. Read up on software development principles and design patterns.

Some more applications outside your PhD thesis would also be a great plus. Consider doing some projects in Natural Language Processing and Computer Vision. I don't want to advertise too much but these two courses are a great starting point:

https://www.udemy.com/course/nlp-natural-language-processing-with-python/

https://www.udemy.com/course/python-for-computer-vision-with-opencv-and-deep-learning/

All this should be very doable in a few months. You could spend the rest of your time studying as much of (advanced) statistics as you can. I feel there is always much to learn and apply. Document your studying with Jupyter Notebooks or create some useful ML models and deploy them in a small scale. Demonstrate that you can write clean, maintainable and scalable code.