• Publications
  • Influence
MeliusNet: Can Binary Neural Networks Achieve MobileNet-level Accuracy?
TLDR
Binary Neural Networks (BNNs) are neural networks which use binary weights and activations instead of the typical 32-bit floating point values. Expand
  • 17
  • 5
  • PDF
Back to Simplicity: How to Train Accurate BNNs from Scratch?
TLDR
We propose a new BNN architecture BinaryDenseNet, which significantly surpasses all existing 1-bit CNNs on ImageNet without tricks. Expand
  • 22
  • 3
  • PDF
BinaryDenseNet: Developing an Architecture for Binary Neural Networks
TLDR
Binary Neural Networks (BNNs) show promising progress in reducing computational and memory costs, but suffer from substantial accuracy degradation compared to their realvalued counterparts on large-scale datasets, e.g., ImageNet. Expand
  • 11
  • 3
  • PDF
Training Competitive Binary Neural Networks from Scratch
TLDR
We present a simple training strategy for binary models without using a pretrained full-precision model. Expand
  • 23
  • 1
  • PDF
Learning to Train a Binary Neural Network
TLDR
We evaluate different network architectures and hyperparameters to provide useful insights on how to train a binary neural network and show how to create efficient models. Expand
  • 8
  • 1
  • PDF
Relative Direction Change - A Topology-based Metric for Layout Stability in Treemaps
TLDR
We present a topology-based metric for layout stability in treemaps—the Relative Direction Change (RDC). Expand
  • 9
  • PDF
Improving Layout Quality by Mixing Treemap-Layouts Based on Data-Change Characteristics
TLDR
This paper presents a hybrid treemap layout approach that optimizes layout-quality metrics by combining state-of-the-art treemAP layout algorithms. Expand
  • 4
KISS: Keeping It Simple for Scene Text Recognition
TLDR
We introduce a new model for scene text recognition that only consists of off-the-shelf building blocks for neural networks. Expand
  • 5
  • PDF
Training Accurate Binary Neural Networks from Scratch
TLDR
Binary neural networks are a promising approach to execute convolutional neural networks on devices with low computational power. Expand
  • 3
BMXNet 2: An Open Source Framework for Low-bit Networks - Reproducing, Understanding, Designing and Showcasing
TLDR
Binary and quantized neural networks are a promising technique to run convolutional neural networks on mobile or embedded devices. Expand