We propose a hybrid architecture composed of a fully convolutional network (FCN) and a Dempster-Shafer layer for image semantic segmentation. In the so-called evidential FCN (E-FCN), an encoder-decoder architecture first extracts pixel-wise feature maps from an input image. A Dempster-Shafer layer then computes mass functions at each pixel location based on distances to prototypes. Finally, a utility layer performs semantic segmentation from mass functions and allows for imprecise classification of ambiguous pixels and outliers. We propose an end-to-end learning strategy for jointly updating the network parameters, which can make use of soft (imprecise) labels. Experiments using three databases (Pascal VOC 2011, MIT-scene Parsing and SIFT Flow) show that the proposed combination improves the accuracy and calibration of semantic segmentation by assigning confusing pixels to multi-class sets.
Ongoing advancements in the fields of 3D modelling and digital archiving have led to an outburst in the amount of data stored digitally. Consequently, several retrieval systems have been developed depending on the type of data stored in these databases. However, unlike text data or images, performing a search for 3D models is non-trivial. Among 3D models, retrieving 3D Engineering/CAD models or mechanical components is even more challenging due to the presence of holes, volumetric features, presence of sharp edges etc., which make CAD a domain unto itself. The research work presented in this paper aims at developing a dataset suitable for building a retrieval system for 3D CAD models based on deep learning. 3D CAD models from the available CAD databases are collected, and a dataset of computer-generated sketch data, termed 'CADSketchNet', has been prepared. Additionally, hand-drawn sketches of the components are also added to CADSketchNet. Using the sketch images from this dataset, the paper also aims at evaluating the performance of various retrieval system or a search engine for 3D CAD models that accepts a sketch image as the input query. Many experimental models are constructed and tested on CADSketchNet. These experiments, along with the model architecture, choice of similarity metrics are reported along with the search results.
Accompanied by the successful progress of deep representation learning, convolutional neural networks (CNNs) have been widely applied to improve the accuracy of polarimetric synthetic aperture radar (PolSAR) image classification. However, in most applications, the difference between PolSAR image and optical image is rarely considered. The design of most existing network structures is not tailored to the characteristics of PolSAR image data and complex-valued data of PolSAR image are simply equated to real-valued data to adapt to the existing mainstream network pipeline to avoid complex-valued operations. These make CNNs unable to perform their full capabilities in the PolSAR image classification tasks. In this paper, we focus on finding a better input form of PolSAR image data and designing special CNN structures that are more compatible with PolSAR image. Considering the relationship between complex number and its amplitude and phase, we extract the amplitude and phase of the complex-valued PolSAR image data as input to maintain the integrity of the original information while avoiding the current immature complex-valued operations, and a novel multi-task CNN framework is proposed to adapt to novel form of input data. Furthermore, in order to better explore the unique phase information in the PolSAR image data, depthwise separable convolutions are applied to the proposed multi-task CNN model. Experiments on three benchmark datasets not only prove that using amplitude and phase information as input does contribute to the improvement of classification accuracy, but also verify the effectiveness of the proposed methods for amplitude and phase input.
Attention modules connecting encoder and decoders have been widely applied in the field of object recognition, image captioning, visual question answering and neural machine translation, and significantly improves the performance. In this paper, we propose a bottom-up gated hierarchical attention (GHA) mechanism for image captioning. Our proposed model employs a CNN as the decoder which is able to learn different concepts at different layers, and apparently, different concepts correspond to different areas of an image. Therefore, we develop the GHA in which low-level concepts are merged into high-level concepts and simultaneously low-level attended features pass to the top to make predictions. Our GHA significantly improves the performance of the model that only applies one level attention, for example, the CIDEr score increases from 0.923 to 0.999, which is comparable to the state-of-the-art models that employ attributes boosting and reinforcement learning (RL). We also conduct extensive experiments to analyze the CNN decoder and our proposed GHA, and we find that deeper decoders cannot obtain better performance, and when the convolutional decoder becomes deeper the model is likely to collapse during training.
In this paper, we aim to improve the performance of semantic image segmentation in a semi-supervised setting in which training is effectuated with a reduced set of annotated images and additional non-annotated images. We present a method based on an ensemble of deep segmentation models. Each model is trained on a subset of the annotated data, and uses the non-annotated images to exchange information with the other models, similar to co-training. Even if each model learns on the same non-annotated images, diversity is preserved with the use of adversarial samples. Our results show that this ability to simultaneously train models, which exchange knowledge while preserving diversity, leads to state-of-the-art results on two challenging medical image datasets.
We have developed a nonlocal algorithm for estimating polarimetric synthetic aperture radar (PolSAR) covariance matrices on single-look complex (SLC) format resolution. The algorithm is inspired by recent work with guided nonlocal means (NLM) speckle filtering, where a co-registered optical image is used to aid the filtering. Based on patch-wise dissimilarities in the SAR and optical domains we set the weights used for the nonlocal average of the outer product of the lexicographic target vectors which form the estimate. By use of this method we show that the estimated covariance matrices preserve the local structure better than previous filtering methods and improve the separation of live from defoliated and dead forest. The detail preserving nature of the algorithm also means that it can be applicable in other settings where preserving the SLC format resolution is necessary.
Fruit flies are established model systems for studying olfactory learning as they will readily learn to associate odors with both electric shock or sugar rewards. The mechanisms of the insect brain apparently responsible for odor learning form a relatively shallow neuronal architecture. Olfactory inputs are received by the antennal lobe (AL) of the brain, which produces an encoding of each odor mixture across ~50 sub-units known as glomeruli. Each of these glomeruli then project its component of this feature vector to several of ~2000 so-called Kenyon Cells (KCs) in a region of the brain known as the mushroom body (MB). Fly responses to odors are generated by small downstream neuropils that decode the higher-order representation from the MB. Research has shown that there is no recognizable pattern in the glomeruli--KC connections (and thus the particular higher-order representations); they are akin to fingerprints~-- even isogenic flies have different projections. Leveraging insights from this architecture, we propose KCNet, a single-hidden-layer neural network that contains sparse, randomized, binary weights between the input layer and the hidden layer and analytically learned weights between the hidden layer and the output layer. Furthermore, we also propose a dynamic optimization algorithm that enables the KCNet to increase performance beyond its structural limits by searching a more efficient set of inputs. For odorant-perception tasks that predict perceptual properties of an odorant, we show that KCNet outperforms existing data-driven approaches, such as XGBoost. For image-classification tasks, KCNet achieves reasonable performance on benchmark datasets (MNIST, Fashion-MNIST, and EMNIST) without any data-augmentation methods or convolutional layers and shows particularly fast running time. Thus, neural networks inspired by the insect brain can be both economical and perform well.
Raven's Progressive Matrices (RPMs) are frequently-used in testing human's visual reasoning ability. Recently developed RPM-like datasets and solution models transfer this kind of problems from cognitive science to computer science. In view of the poor generalization performance due to insufficient samples in RPM datasets, we propose a data augmentation strategy by image mix-up, which is generalizable to a variety of multiple-choice problems, especially for image-based RPM-like problems. By focusing on potential functionalities of negative candidate answers, the visual reasoning capability of the model is enhanced. By applying the proposed data augmentation method, we achieve significant and consistent improvement on various RPM-like datasets compared with the state-of-the-art models.
Low dose computed tomography (LDCT) has attracted more and more attention in routine clinical diagnosis assessment, therapy planning, etc., which can reduce the dose of X-ray radiation to patients. However, the noise caused by low X-ray exposure degrades the CT image quality and then affects clinical diagnosis accuracy. In this paper, we train a transformer-based neural network to enhance the final CT image quality. To be specific, we first decompose the noisy LDCT image into two parts: high-frequency (HF) and low-frequency (LF) compositions. Then, we extract content features (X_{L_c}) and latent texture features (X_{L_t}) from the LF part, as well as HF embeddings (X_{H_f}) from the HF part. Further, we feed X_{L_t} and X_{H_f} into a modified transformer with three encoders and decoders to obtain well-refined HF texture features. After that, we combine these well-refined HF texture features with the pre-extracted X_{L_c} to encourage the restoration of high-quality LDCT images with the assistance of piecewise reconstruction. Extensive experiments on Mayo LDCT dataset show that our method produces superior results and outperforms other methods.
Given an image of a target person and an image of another person wearing a garment, we automatically generate the target person in the given garment. At the core of our method is a pose-conditioned StyleGAN2 latent space interpolation, which seamlessly combines the areas of interest from each image, i.e., body shape, hair, and skin color are derived from the target person, while the garment with its folds, material properties, and shape comes from the garment image. By automatically optimizing for interpolation coefficients per layer in the latent space, we can perform a seamless, yet true to source, merging of the garment and target person. Our algorithm allows for garments to deform according to the given body shape, while preserving pattern and material details. Experiments demonstrate state-of-the-art photo-realistic results at high resolution ($512\times 512$).