Accelerating Deep Learning Frameworks with Micro-Batches | IEEE Conference Publication | IEEE Xplore