The transformer networks, which originate from machine translation, are particularly good at modeling long-range dependencies within a long sequence. Currently, the transformer networks are making revolutionary progress in various vision tasks ranging from high-level classification tasks to low-level dense prediction tasks. In this paper, we conduct research on applying the transformer networks for salient object detection (SOD). Specifically, we adopt the dense transformer backbone for fully supervised RGB image based SOD, RGB-D image pair based SOD, and weakly supervised SOD via scribble supervision. As an extension, we also apply our fully supervised model to the task of camouflaged object detection (COD) for camouflaged object segmentation. For the fully supervised models, we define the dense transformer backbone as feature encoder, and design a very simple decoder to produce a one channel saliency map (or camouflage map for the COD task). For the weakly supervised model, as there exists no structure information in the scribble annotation, we first adopt the recent proposed Gated-CRF loss to effectively model the pair-wise relationships for accurate model prediction. Then, we introduce self-supervised learning strategy to push the model to produce scale-invariant predictions, which is proven effective for weakly supervised models and models trained on small training datasets. Extensive experimental results on various SOD and COD tasks (fully supervised RGB image based SOD, fully supervised RGB-D image pair based SOD, weakly supervised SOD via scribble supervision, and fully supervised RGB image based COD) illustrate that transformer networks can transform salient object detection and camouflaged object detection, leading to new benchmarks for each related task.
Despite the remarkable success, deep learning models have shown to be vulnerable to the universal adversarial perturbation (UAP). The existing endeavors on UAP methods mainly focus on attacking the image classification models. Nevertheless, little attention has been paid to attacking image retrieval systems. In this paper, we make the first attempt for UAP attacking to deep feature based image retrieval. Concretely, attacking image retrieval is to make the retrieval system return more irrelevant images to the query at the top ranking list, whose key design is to corrupt the relationships among features. To this end, we propose a unified method to generate retrieval-based UAP to break the relationships between image features from point-wise, label-wise, and list-wise aspects. We further analyze the impact of the resizing operation in generating UAP, and thus provide a solution to attack high-performance retrieval systems with query resizing. We evaluate the proposed methods on four widely-used image retrieval datasets, i.e., Oxford5k and Paris6k with their revised versions, which lead to a significant performance drop in terms of different metrics, such as mAP, and mP@10. Finally, we test our attacking methods on the real-world visual search engine, i.e., Google Images, which demonstrates the potential of our methods.
Computer-aided analysis of biological images typically requires extensive training on large-scale annotated datasets, which is not viable in many situations. In this paper we present GAN-DL, a Discriminator Learner based on the StyleGAN2 architecture, which we employ for self-supervised image representation learning in the case of fluorescent biological images. We show that Wasserstein Generative Adversarial Networks combined with linear Support Vector Machines enable high-throughput compound screening based on raw images. We demonstrate this by classifying active and inactive compounds tested for the inhibition of SARS-CoV-2 infection in VERO and HRCE cell lines. In contrast to previous methods, our deep learning based approach does not require any annotation besides the one that is normally collected during the sample preparation process. We test our technique on the RxRx19a Sars-CoV-2 image collection. The dataset consists of fluorescent images that were generated to assess the ability of regulatory-approved or in late-stage clinical trials compound to modulate the in vitro infection from SARS-CoV-2 in both VERO and HRCE cell lines. We show that our technique can be exploited not only for classification tasks, but also to effectively derive a dose response curve for the tested treatments, in a self-supervised manner. Lastly, we demonstrate its generalization capabilities by successfully addressing a zero-shot learning task, consisting in the categorization of four different cell types of the RxRx1 fluorescent images collection.
Recently proposed few-shot image classification methods have generally focused on use cases where the objects to be classified are the central subject of images. Despite success on benchmark vision datasets aligned with this use case, these methods typically fail on use cases involving densely-annotated, busy images: images common in the wild where objects of relevance are not the central subject, instead appearing potentially occluded, small, or among other incidental objects belonging to other classes of potential interest. To localize relevant objects, we employ a prototype-based few-shot segmentation model which compares the encoded features of unlabeled query images with support class centroids to produce region proposals indicating the presence and location of support set classes in a query image. These region proposals are then used as additional conditioning input to few-shot image classifiers. We develop a framework to unify the two stages (segmentation and classification) into an end-to-end classification model -- PRoPnet -- and empirically demonstrate that our methods improve accuracy on image datasets with natural scenes containing multiple object classes.
In this thesis, we offer a thorough investigation of different regularisation terms used in variational imaging problems, together with detailed optimisation processes of these problems. We begin by studying smooth problems and partially non-smooth problems in the form of Tikhonov denoising and Total Variation (TV) denoising, respectively. For Tikhonov denoising, we study an accelerated gradient method with adaptive restart, which shows a very rapid convergence rate. However, it is not straightforward to apply this fast algorithm to TV denoising, due to the non-smoothness of its built-in regularisation. To tackle this issue, we propose to utilise duality to convert such a non-smooth problem into a smooth one so that the accelerated gradient method with restart applies naturally. However, we notice that both Tikhonov and TV regularisations have drawbacks, in the form of blurred image edges and staircase artefacts, respectively. To overcome these drawbacks, we propose a novel adaption to Total Generalised Variation (TGV) regularisation called Total Smooth Variation (TSV), which retains edges and meanwhile does not produce results which contain staircase artefacts. To optimise TSV effectively, we then propose the Accelerated Proximal Gradient Algorithm (APGA) which also utilises adaptive restart techniques. Compared to existing state-of-the-art regularisations (e.g. TV), TSV is shown to obtain more effective results on denoising problems as well as advanced imaging applications such as magnetic resonance imaging (MRI) reconstruction and optical flow. TSV removes the staircase artefacts observed when using TV regularisation, but has the added advantage over TGV that it can be efficiently optimised using gradient based methods with Nesterov acceleration and adaptive restart. Code is available at https://github.com/Jbartlett6/Accelerated-First-Order-Method-for-Variational-Imaging.
Camouflaged objects attempt to conceal their texture into the background and discriminating them from the background is hard even for human beings. The main objective of this paper is to explore the camouflaged object segmentation problem, namely, segmenting the camouflaged object(s) for a given image. This problem has not been well studied in spite of a wide range of potential applications including the preservation of wild animals and the discovery of new species, surveillance systems, search-and-rescue missions in the event of natural disasters such as earthquakes, floods or hurricanes. This paper addresses a new challenging problem of camouflaged object segmentation. To address this problem, we provide a new image dataset of camouflaged objects for benchmarking purposes. In addition, we propose a general end-to-end network, called the Anabranch Network, that leverages both classification and segmentation tasks. Different from existing networks for segmentation, our proposed network possesses the second branch for classification to predict the probability of containing camouflaged object(s) in an image, which is then fused into the main branch for segmentation to boost up the segmentation accuracy. Extensive experiments conducted on the newly built dataset demonstrate the effectiveness of our network using various fully convolutional networks. \url{https://sites.google.com/view/ltnghia/research/camo}
Hyperspectral images (HSIs) are unavoidably corrupted by mixed noise which hinders the subsequent applications. Traditional methods exploit the structure of the HSI via optimization-based models for denoising, while their capacity is inferior to the convolutional neural network (CNN)-based methods, which supervisedly learn the noisy-to-denoised mapping from a large amount of data. However, as the clean-noisy pairs of hyperspectral data are always unavailable in many applications, it is eager to build an unsupervised HSI denoising method with high model capability. To remove the mixed noise in HSIs, we suggest the spatial-spectral constrained deep image prior (S2DIP), which simultaneously capitalize the high model representation ability brought by the CNN in an unsupervised manner and does not need any extra training data. Specifically, we employ the separable 3D convolution blocks to faithfully encode the HSI in the framework of DIP, and a spatial-spectral total variation (SSTV) term is tailored to explore the spatial-spectral smoothness of HSIs. Moreover, our method favorably addresses the semi-convergence behavior of prevailing unsupervised methods, e.g., DIP 2D, and DIP 3D. Extensive experiments demonstrate that the proposed method outperforms state-of-the-art optimization-based HSI denoising methods in terms of effectiveness and robustness.
Trained with a sufficiently large training and testing dataset, Deep Neural Networks (DNNs) are expected to generalize. However, inputs may deviate from the training dataset distribution in real deployments. This is a fundamental issue with using a finite dataset. Even worse, real inputs may change over time from the expected distribution. Taken together, these issues may lead deployed DNNs to mis-predict in production. In this work, we present a runtime approach that mitigates DNN mis-predictions caused by the unexpected runtime inputs to the DNN. In contrast to previous work that considers the structure and parameters of the DNN itself, our approach treats the DNN as a blackbox and focuses on the inputs to the DNN. Our approach has two steps. First, it recognizes and distinguishes "unseen" semantically-preserving inputs. For this we use a distribution analyzer based on the distance metric learned by a Siamese network. Second, our approach transforms those unexpected inputs into inputs from the training set that are identified as having similar semantics. We call this process input reflection and formulate it as a search problem over the embedding space on the training set. This embedding space is learned by a Quadruplet network as an auxiliary model for the subject model to improve the generalization. We implemented a tool called InputReflector based on the above two-step approach and evaluated it with experiments on three DNN models trained on CIFAR-10, MNIST, and FMINST image datasets. The results show that InputReflector can effectively distinguish inputs that retain semantics of the distribution (e.g., blurred, brightened, contrasted, and zoomed images) and out-of-distribution inputs from normal inputs.
Along with predictive performance and runtime speed, reliability is a key requirement for real-world semantic segmentation. Reliability encompasses robustness, predictive uncertainty and reduced bias. To improve reliability, we introduce Superpixel-mix, a new superpixel-based data augmentation method with teacher-student consistency training. Unlike other mixing-based augmentation techniques, mixing superpixels between images is aware of object boundaries, while yielding consistent gains in segmentation accuracy. Our proposed technique achieves state-of-the-art results in semi-supervised semantic segmentation on the Cityscapes dataset. Moreover, Superpixel-mix improves the reliability of semantic segmentation by reducing network uncertainty and bias, as confirmed by competitive results under strong distributions shift (adverse weather, image corruptions) and when facing out-of-distribution data.
Indoor image features extraction is a fundamental problem in multiple fields such as image processing, pattern recognition, robotics and so on. Nevertheless, most of the existing feature extraction methods, which extract features based on pixels, color, shape/object parts or objects on images, suffer from limited capabilities in describing semantic information (e.g., object association). These techniques, therefore, involve undesired classification performance. To tackle this issue, we propose the notion of high-level semantic features and design four steps to extract them. Specifically, we first construct the objects pattern dictionary through extracting raw objects in the images, and then retrieve and extract semantic objects from the objects pattern dictionary. We finally extract our high-level semantic features based on the calculated probability and delta parameter. Experiments on three publicly available datasets (MIT-67, Scene15 and NYU V1) show that our feature extraction approach outperforms state-of-the-art feature extraction methods for indoor image classification, given a lower dimension of our features than those methods.