Segmentation of ultra-high resolution images is challenging because of their enormous size, consisting of millions or even billions of pixels. Typical solutions include dividing input images into patches of fixed size and/or down-sampling to meet memory constraints. Such operations incur information loss in the field-of-view (FoV) i.e., spatial coverage and the image resolution. The impact on segmentation performance is, however, as yet understudied. In this work, we start with a motivational experiment which demonstrates that the trade-off between FoV and resolution affects the segmentation performance on ultra-high resolution images---and furthermore, its influence also varies spatially according to the local patterns in different areas. We then introduce foveation module, a learnable "dataloader" which, for a given ultra-high resolution image, adaptively chooses the appropriate configuration (FoV/resolution trade-off) of the input patch to feed to the downstream segmentation model at each spatial location of the image. The foveation module is jointly trained with the segmentation network to maximise the task performance. We demonstrate on three publicly available high-resolution image datasets that the foveation module consistently improves segmentation performance over the cases trained with patches of fixed FoV/resolution trade-off. Our approach achieves the SoTA performance on the DeepGlobe aerial image dataset. On the Gleason2019 histopathology dataset, our model achieves better segmentation accuracy for the two most clinically important and ambiguous classes (Gleason Grade 3 and 4) than the top performers in the challenge by 13.1% and 7.5%, and improves on the average performance of 6 human experts by 6.5% and 7.5%. Our code and trained models are available at $\text{https://github.com/lxasqjc/Foveation-Segmentation}$.
GANs (Generative adversarial networks) is a new AI technology that can perform deep learning with less training data and has the capability of achieving transformation between two image sets. Using GAN we have carried out a comparison between several art sets with different art style. We have prepared several image sets; a flower photo set (A), an art image set (B1) of Impressionism drawings, an art image set of abstract paintings (B2), an art image set of Chinese figurative paintings, (B3), and an art image set of abstract images (B4) created by Naoko Tosa, one of the authors. Transformation between set A to each of B was carried out using GAN and four image sets (B1, B2, B3, B4) was obtained. Using these four image sets we have carried out psychological experiment by asking subjects consisting of 23 students to fill in questionnaires. By analyzing the obtained questionnaires, we have found the followings. Abstract drawings and figurative drawings are clearly judged to be different. Figurative drawings in West and East were judged to be similar. Abstract images by Naoko Tosa were judged as similar to Western abstract images. These results show that AI could be used as an analysis tool to reveal differences between art genres.
Noise injection-based regularization, such as Dropout, has been widely used in image domain to improve the performance of deep neural networks (DNNs). However, efficient regularization in the point cloud domain is rarely exploited, and most of the state-of-the-art works focus on data augmentation-based regularization. In this paper, we, for the first time, perform systematic investigation on noise injection-based regularization for point cloud-domain DNNs. To be specific, we propose a series of regularization techniques, namely DropFeat, DropPoint and DropCluster, to perform noise injection on the point feature maps at the feature level, point level and cluster level, respectively. We also empirically analyze the impacts of different factors, including dropping rate, cluster size and dropping position, to obtain useful insights and general deployment guidelines, which can facilitate the adoption of our approaches across different datasets and DNN architectures. We evaluate our proposed approaches on various DNN models for different point cloud processing tasks. Experimental results show our approaches enable significant performance improvement. Notably, our DropCluster brings 1.5%, 1.3% and 0.8% higher overall accuracy for PointNet, PointNet++ and DGCNN, respectively, on ModelNet40 shape classification dataset. On ShapeNet part segmentation dataset, DropCluster brings 0.5%, 0.5% and 0.2% mean Intersection-over-union (IoU) increase for PointNet, PointNet++ and DGCNN, respectively. On S3DIS semantic segmentation dataset, DropCluster improves the mean IoU of PointNet, PointNet++ and DGCNN by 3.2%, 2.9% and 3.7%, respectively. Meanwhile, DropCluster also enables the overall accuracy increase for these three popular backbone DNNs by 2.4%, 2.2% and 1.8%, respectively.
Video Denoising is one of the fundamental tasks of any videoprocessing pipeline. It is different from image denoising due to the tem-poral aspects of video frames, and any image denoising approach appliedto videos will result in flickering. The proposed method makes use oftemporal as well as spatial dimensions of video frames as part of a two-stage pipeline. Each stage in the architecture named as Spatio-TemporalNetwork uses a channel-wise attention mechanism to forward the encodersignal to the decoder side. The Attention Block used in this paper usessoft attention to ranks the filters for better training.
With the improvement of pattern recognition and feature extraction of Deep Neural Networks (DNNs), more and more problems are attempted to solve from the view of images. Recently, a Reconstructive Neural Network (ReConNN) was proposed to obtain an image-based model from an analysis-based model, which can help us to solve many high frequency problems with difficult sampling, e.g. sonic wave and collision. However, due to the slight difference between simulated images, the low-accuracy of the Convolutional Neural Network (CNN) and poor-diversity of the Generative Adversarial Network (GAN) make the reconstruction process low-accuracy, poor-efficiency, expensive-computation and high-manpower. In this study, an improved ReConNN model is proposed to address the mentioned weaknesses. Through experiments, comparisons and analyses, the improved one is demonstrated to outperform in accuracy, efficiency and cost.
The right to be forgotten states that a data subject has the right to erase their data from an entity storing it. In the context of machine learning (ML), it requires the ML model provider to remove the data subject's data from the training set used to build the ML model, a process known as \textit{machine unlearning}. While straightforward and legitimate, retraining the ML model from scratch upon receiving unlearning requests incurs high computational overhead when the training set is large. To address this issue, a number of approximate algorithms have been proposed in the domain of image and text data, among which SISA is the state-of-the-art solution. It randomly partitions the training set into multiple shards and trains a constituent model for each shard. However, directly applying SISA to the graph data can severely damage the graph structural information, and thereby the resulting ML model utility. In this paper, we propose GraphEraser, a novel machine unlearning method tailored to graph data. Its contributions include two novel graph partition algorithms, and a learning-based aggregation method. We conduct extensive experiments on five real-world datasets to illustrate the unlearning efficiency and model utility of GraphEraser. We observe that GraphEraser achieves 2.06$\times$ (small dataset) to 35.94$\times$ (large dataset) unlearning time improvement compared to retraining from scratch. On the other hand, GraphEraser achieves up to $62.5\%$ higher F1 score than that of random partitioning. In addition, our proposed learning-based aggregation method achieves up to $112\%$ higher F1 score than that of the majority vote aggregation.
Batch Normalization is a key component in almost all state-of-the-art image classifiers, but it also introduces practical challenges: it breaks the independence between training examples within a batch, can incur compute and memory overhead, and often results in unexpected bugs. Building on recent theoretical analyses of deep ResNets at initialization, we propose a simple set of analysis tools to characterize signal propagation on the forward pass, and leverage these tools to design highly performant ResNets without activation normalization layers. Crucial to our success is an adapted version of the recently proposed Weight Standardization. Our analysis tools show how this technique preserves the signal in networks with ReLU or Swish activation functions by ensuring that the per-channel activation means do not grow with depth. Across a range of FLOP budgets, our networks attain performance competitive with the state-of-the-art EfficientNets on ImageNet.
The inability to interpret the model prediction in semantically and visually meaningful ways is a well-known shortcoming of most existing computer-aided diagnosis methods. In this paper, we propose MDNet to establish a direct multimodal mapping between medical images and diagnostic reports that can read images, generate diagnostic reports, retrieve images by symptom descriptions, and visualize attention, to provide justifications of the network diagnosis process. MDNet includes an image model and a language model. The image model is proposed to enhance multi-scale feature ensembles and utilization efficiency. The language model, integrated with our improved attention mechanism, aims to read and explore discriminative image feature descriptions from reports to learn a direct mapping from sentence words to image pixels. The overall network is trained end-to-end by using our developed optimization strategy. Based on a pathology bladder cancer images and its diagnostic reports (BCIDR) dataset, we conduct sufficient experiments to demonstrate that MDNet outperforms comparative baselines. The proposed image model obtains state-of-the-art performance on two CIFAR datasets as well.
Face recognition can benefit from the utilization of depth data captured using low-cost cameras, in particular for presentation attack detection purposes. Depth video output from these capture devices can however contain defects such as holes, as well as general depth inaccuracies. This work proposes a deep learning-based face depth enhancement method. The trained artificial neural networks utilize U-Net-like architectures, and are compared against general enhancer types. All tested enhancer types exclusively use depth data as input, which differs from methods that enhance depth based on additional input data such as visible light color images. Due to the noted apparent lack of real-world camera datasets with suitable properties, face depth ground truth images and degraded forms thereof are synthesized with help of PRNet, both for the deep learning training and for an experimental quantitative evaluation of all enhancer types. Generated enhancer output samples are also presented for real camera data, namely custom RealSense D435 depth images and Kinect v1 data from the KinectFaceDB. It is concluded that the deep learning enhancement approach is superior to the tested general enhancers, without overly falsifying depth data when non-face input is provided.
Recently, there has been a lot of interest in automatically generating descriptions for an image. Most existing language-model based approaches for this task learn to generate an image description word by word in its original word order. However, for humans, it is more natural to locate the objects and their relationships first, and then elaborate on each object, describing notable attributes. We present a coarse-to-fine method that decomposes the original image description into a skeleton sentence and its attributes, and generates the skeleton sentence and attribute phrases separately. By this decomposition, our method can generate more accurate and novel descriptions than the previous state-of-the-art. Experimental results on the MS-COCO and a larger scale Stock3M datasets show that our algorithm yields consistent improvements across different evaluation metrics, especially on the SPICE metric, which has much higher correlation with human ratings than the conventional metrics. Furthermore, our algorithm can generate descriptions with varied length, benefiting from the separate control of the skeleton and attributes. This enables image description generation that better accommodates user preferences.