Simple Fully Convolutional Network Research Articles

Robust residual convolutional neural network based pupil tracking for low-computational power applications

Existing Deep Fully Convolutional Network (DFCN) based pupil segmentation models have been shown to perform robust and accurate pupil detection. However, they lack high computational resources when real-time predictions for low computational wearable devices such as Raspberry Pi are of interest. Simple FCN (Fully Convolutional Network) models, on the other hand, can provide real-time yet not as accurate predictions with low computational power devices. In this study, we address this dilemma by proposing a CNN (Convolutional Neural Network)-based model, Residual CNN, with several advanced operations such as residual connections, Squeeze and Excitation (SE) attention, and Atrous Spatial Pyramid Pooling (ASPP) to improve the prediction performance without complicating the model. Moreover, we leverage transfer learning by training synthetic images, then fine-tuning them with authentic eye images. We also fully quantize our model parameters to speed up the predictions and apply the Quantization Aware Training (QAT) strategy to obtain accurate predictions. Our experiments show that our quantized Res-CNN model trained by QAT strategy with 40x30 resolution images provides robust yet real-time predictions with an average of 1.351 Root Mean Square Error (RMSE) for the pupil center predictions and 8.317 ms response time per image on Raspberry Pi.

Convolutional neural network with median layers for denoising salt-and-pepper contaminations

We propose a deep fully convolutional neural network with a new type of layer, named median layer, to restore images contaminated by salt-and-pepper (s&p) noise. A median layer simply performs median filtering on all feature channels. By adding this kind of layer into some widely used fully convolutional deep neural networks, we develop an end-to-end network that removes extremely high-level s&p noise without performing any non-trivial preprocessing tasks. Experiments show that inserting median layers into a simple fully-convolutional network with the L2 loss significantly boosts signal-to-noise ratio. Quantitative comparisons testify that our network outperforms the state-of-the-art methods with a limited amount of training data.

Accurate brain age prediction with lightweight deep neural networks.

Deep learning has huge potential for accurate disease prediction with neuroimaging data, but the prediction performance is often limited by training-dataset size and computing memory requirements. To address this, we propose a deep convolutional neural network model, Simple Fully Convolutional Network (SFCN), for accurate prediction of brain age using T1-weighted structural MRI data. Compared with other popular deep network architectures, SFCN has fewer parameters, so is more compatible with small dataset size and 3D volume data. The network architecture was combined with several techniques for boosting performance, including data augmentation, pre-training, model regularization, model ensemble and prediction bias correction. We compared our overall SFCN approach with several widely-used machine learning models. It achieved state-of-the-art performance in UK Biobank data (N = 14,503), with mean absolute error (MAE) = 2.14y in brain age prediction and 99.5% in sex classification. SFCN also won (both parts of) the 2019 Predictive Analysis Challenge for brain age prediction, involving 79 competing teams (N = 2,638, MAE = 2.90y). We describe here the details of our approach, and its optimisation and validation. Our approach can easily be generalised to other tasks using different image modalities, and is released on GitHub.

WideSegNeXt: Semantic Image Segmentation Using Wide Residual Network and NeXt Dilated Unit

Semantic segmentation is widely applied in autonomous driving, in robotic picking, and for medical purposes. Due to the breakthrough of deep learning in recent years, the fully convolutional network (FCN)-based method has become the de facto standard in semantic segmentation. However, the simple FCN has difficulty in capturing global context information, since the local receptive field is small. Furthermore, there is a problem of low image resolution because of the existence of the pooling layer. In this paper, we address the shortcomings of the FCN by proposing a new architecture called WideSegNeXt, which captures the image context on various spatial scales and is effective in identifying small objects. In addition, there is little loss of position information, since there are no pooling layers in the structure. The proposed method achieves a mean intersection over union (MIoU) of 72.5% and a global accuracy (GA) of 92.4% on the CamVid dataset and achieves higher performance than previous methods without additional input datasets.