ItNet: iterative neural networks for fast and efficient anytime prediction

Thomas Pfeil

Keywords: efficient deep neural network, semantic segmentation, parameter sharing, anytime prediction, tiny network graph, massively parallel hardware systems, recurrent convolutional network

Abstract: Deep neural networks have usually to be compressed and accelerated for their usage in low-power, e.g. mobile, devices. Common requirements are high accuracy, high throughput, low latency, and a small memory footprint. A good trade-off between accuracy and latency has been shown by networks comprising multiple intermediate outputs. In this study, we introduce a multi-output network that has a tiny memory footprint in terms of its computational graph, which allows its execution on novel, massively-parallel hardware accelerators designed for extremely high throughput. To this end, the graph is designed to contain loops by iteratively executing a single network building block. These so-called iterative neural networks enable stateof-the-art results for semantic segmentation on the CamVid and Cityscapes datasets that are especially demanding in terms of computational resources. In ablation studies, the improvement of network training by intermediate network outputs as well as the trade-off between weight sharing over iterations and the network size are investigated.

One-sentence Summary: We investigate the trade-off between the prediction accuracy and the size of the computational graph to allow the neural network to be executed on massively-parallel hardware systems.

Code Of Ethics: I acknowledge that I and all co-authors of this work have read and commit to adhering to the ICLR Code of Ethics

Reviewed Version (pdf): https://openreview.net/references/pdf?id=ZcH97EbSeQ