I am currently trying to train medical images with multiple (around 30) different objects. I want to train my image data for object detection, which means from a single image it should detect 30 objects separately.
I have two GTX 1080 GPUs. But, I cannot train more than 7 images (with multiple objects). My image size is around 1500 by 1000. It is showing out of memory. I am using MATLAB 2018b.
My supervisor is suggesting me to change the platform to Python (Pytorch/tensorflow). I want to know whether it really matters that I am using MATLAB or Python for training? Does Python offer more favorable GPU performance than MATLAB?
Dear Sir,
I want to ask a simple question.
I am currently trying to train medical images with multiple (around 30) different objects. I want to train my image data for object detection, which means from a single image it should detect 30 objects separately.
I have two GTX 1080 GPUs. But, I cannot train more than 7 images (with multiple objects). My image size is around 1500 by 1000. It is showing out of memory. I am using MATLAB 2018b.
My supervisor is suggesting me to change the platform to Python (Pytorch/tensorflow). I want to know whether it really matters that I am using MATLAB or Python for training? Does Python offer more favorable GPU performance than MATLAB?
Please clarify.