Unsupervised learning is just at a tipping point where it could really take off. Among these approaches, contrastive learning has seen tremendous progress and led to state-of-the-art performance. In this paper, we construct a novel probabilistic graphical model that effectively incorporates the low rank promoting prior into the framework of contrastive learning, referred to as LORAC. In contrast to the existing conventional self-supervised approaches that only considers independent learning, our hypothesis explicitly requires that all the samples belonging to the same instance class lie on the same subspace with small dimension. This heuristic poses particular joint learning constraints to reduce the degree of freedom of the problem during the search of the optimal network parameterization. Most importantly, we argue that the low rank prior employed here is not unique, and many different priors can be invoked in a similar probabilistic way, corresponding to different hypotheses about underlying truth behind the contrastive features. Empirical evidences show that the proposed algorithm clearly surpasses the state-of-the-art approaches on multiple benchmarks, including image classification, object detection, instance segmentation and keypoint detection.
Software testing is often hindered where it is impossible or impractical to determine the correctness of the behaviour or output of the software under test (SUT), a situation known as the oracle problem. An example of an area facing the oracle problem is automatic image classification, using machine learning to classify an input image as one of a set of predefined classes. An approach to software testing that alleviates the oracle problem is metamorphic testing (MT). While traditional software testing examines the correctness of individual test cases, MT instead examines the relations amongst multiple executions of test cases and their outputs. These relations are called metamorphic relations (MRs): if an MR is found to be violated, then a fault must exist in the SUT. This paper examines the problem of classifying images containing visually hidden markers called Artcodes, and applies MT to verify and enhance the trained classifiers. This paper further examines two MRs, Separation and Occlusion, and reports on their capability in verifying the image classification using one-way analysis of variance (ANOVA) in conjunction with three other statistical analysis methods: t-test (for unequal variances), Kruskal-Wallis test, and Dunnett's test. In addition to our previously-studied classifier, that used Random Forests, we introduce a new classifier that uses a support vector machine, and present its MR-augmented version. Experimental evaluations across a number of performance metrics show that the augmented classifiers can achieve better performance than non-augmented classifiers. This paper also analyses how the enhanced performance is obtained.
The nonlocal-based blocks are designed for capturing long-range spatial-temporal dependencies in computer vision tasks. Although having shown excellent performance, they still lack the mechanism to encode the rich, structured information among elements in an image or video. In this paper, to theoretically analyze the property of these nonlocal-based blocks, we provide a new perspective to interpret them, where we view them as a set of graph filters generated on a fully-connected graph. Specifically, when choosing the Chebyshev graph filter, a unified formulation can be derived for explaining and analyzing the existing nonlocal-based blocks (e.g., nonlocal block, nonlocal stage, double attention block). Furthermore, by concerning the property of spectral, we propose an efficient and robust spectral nonlocal block, which can be more robust and flexible to catch long-range dependencies when inserted into deep neural networks than the existing nonlocal blocks. Experimental results demonstrate the clear-cut improvements and practical applicabilities of our method on image classification, action recognition, semantic segmentation, and person re-identification tasks.
Deep learning has achieved great success in a wide spectrum of multimedia applications such as image classification, natural language processing and multimodal data analysis. Recent years have seen the development of many deep learning frameworks that provide a high-level programming interface for users to design models, conduct training and deploy inference. However, it remains challenging to build an efficient end-to-end multimedia application with most existing frameworks. Specifically, in terms of usability, it is demanding for non-experts to implement deep learning models, obtain the right settings for the entire machine learning pipeline, manage models and datasets, and exploit external data sources all together. Further, in terms of adaptability, elastic computation solutions are much needed as the actual serving workload fluctuates constantly, and scaling the hardware resources to handle the fluctuating workload is typically infeasible. To address these challenges, we introduce SINGA-Easy, a new deep learning framework that provides distributed hyper-parameter tuning at the training stage, dynamic computational cost control at the inference stage, and intuitive user interactions with multimedia contents facilitated by model explanation. Our experiments on the training and deployment of multi-modality data analysis applications show that the framework is both usable and adaptable to dynamic inference loads. We implement SINGA-Easy on top of Apache SINGA and demonstrate our system with the entire machine learning life cycle.
In this paper, we propose neural networks that tackle the problems of stability and field-of-view of a Convolutional Neural Network (CNN). As an alternative to increasing the network's depth or width to improve performance, we propose integral-based spatially nonlocal operators which are related to global weighted Laplacian, fractional Laplacian and inverse fractional Laplacian operators that arise in several problems in the physical sciences. The forward propagation of such networks is inspired by partial integro-differential equations (PIDEs). We test the effectiveness of the proposed neural architectures on benchmark image classification datasets and semantic segmentation tasks in autonomous driving. Moreover, we investigate the extra computational costs of these dense operators and the stability of forward propagation of the proposed neural networks.
Most existing neural architecture search (NAS) algorithms are dedicated to and evaluated by the downstream tasks, e.g., image classification in computer vision. However, extensive experiments have shown that, prominent neural architectures, such as ResNet in computer vision and LSTM in natural language processing, are generally good at extracting patterns from the input data and perform well on different downstream tasks. In this paper, we attempt to answer two fundamental questions related to NAS. (1) Is it necessary to use the performance of specific downstream tasks to evaluate and search for good neural architectures? (2) Can we perform NAS effectively and efficiently while being agnostic to the downstream tasks? To answer these questions, we propose a novel and generic NAS framework, termed Generic NAS (GenNAS). GenNAS does not use task-specific labels but instead adopts regression on a set of manually designed synthetic signal bases for architecture evaluation. Such a self-supervised regression task can effectively evaluate the intrinsic power of an architecture to capture and transform the input signal patterns, and allow more sufficient usage of training samples. Extensive experiments across 13 CNN search spaces and one NLP space demonstrate the remarkable efficiency of GenNAS using regression, in terms of both evaluating the neural architectures (quantified by the ranking correlation Spearman's rho between the approximated performances and the downstream task performances) and the convergence speed for training (within a few seconds).
Transformers with powerful global relation modeling abilities have been introduced to fundamental computer vision tasks recently. As a typical example, the Vision Transformer (ViT) directly applies a pure transformer architecture on image classification, by simply splitting images into tokens with a fixed length, and employing transformers to learn relations between these tokens. However, such naive tokenization could destruct object structures, assign grids to uninterested regions such as background, and introduce interference signals. To mitigate the above issues, in this paper, we propose an iterative and progressive sampling strategy to locate discriminative regions. At each iteration, embeddings of the current sampling step are fed into a transformer encoder layer, and a group of sampling offsets is predicted to update the sampling locations for the next step. The progressive sampling is differentiable. When combined with the Vision Transformer, the obtained PS-ViT network can adaptively learn where to look. The proposed PS-ViT is both effective and efficient. When trained from scratch on ImageNet, PS-ViT performs 3.8% higher than the vanilla ViT in terms of top-1 accuracy with about $4\times$ fewer parameters and $10\times$ fewer FLOPs. Code is available at https://github.com/yuexy/PS-ViT.
The application of artificial intelligence (AI) and data-driven decision-making systems in autonomous vehicles is growing rapidly. As autonomous vehicles operate in dynamic environments, the risk that they can face an unknown observation is relatively high due to insufficient training data, distributional shift, or cyber-security attack. Thus, AI-based algorithms should make dependable decisions to improve their interpretation of the environment, lower the risk of autonomous driving, and avoid catastrophic accidents. This paper proposes an approach named SafeML II, which applies empirical cumulative distribution function (ECDF)-based statistical distance measures in a designed human-in-the-loop procedure to ensure the safety of machine learning-based classifiers in autonomous vehicle software. The approach is model-agnostic and it can cover various machine learning and deep learning classifiers. The German Traffic Sign Recognition Benchmark (GTSRB) is used to illustrate the capabilities of the proposed approach.
In many real-world applications, we often need to handle various deployment scenarios, where the resource constraint and the superclass of interest corresponding to a group of classes are dynamically specified. How to efficiently deploy deep models for diverse deployment scenarios is a new challenge. Previous NAS approaches seek to design architectures for all classes simultaneously, which may not be optimal for some individual superclasses. A straightforward solution is to search an architecture from scratch for each deployment scenario, which however is computation-intensive and impractical. To address this, we present a novel and general framework, called Elastic Architecture Search (EAS), permitting instant specializations at runtime for diverse superclasses with various resource constraints. To this end, we first propose to effectively train an over-parameterized network via a superclass dropout strategy during training. In this way, the resulting model is robust to the subsequent superclasses dropping at inference time. Based on the well-trained over-parameterized network, we then propose an efficient architecture generator to obtain promising architectures within a single forward pass. Experiments on three image classification datasets show that EAS is able to find more compact networks with better performance while remarkably being orders of magnitude faster than state-of-the-art NAS methods, e.g., outperforming OFA (once-for-all) by 1.3% on Top-1 accuracy at a budget around 361M #MAdds on ImageNet-10. More critically, EAS is able to find compact architectures within 0.1 second for 50 deployment scenarios.
Image classification must work for autonomous vehicles (AV) operating on public roads, and actions performed based on image misclassification can have serious consequences. Traffic sign images can be misclassified by an adversarial attack on machine learning models used by AVs for traffic sign recognition. To make classification models resilient against adversarial attacks, we used a hybrid deep-learning model with both the quantum and classical layers. Our goal is to study the hybrid deep-learning architecture for classical-quantum transfer learning models to support the current era of intermediate-scale quantum technology. We have evaluated the impacts of various white box adversarial attacks on these hybrid models. The classical part of hybrid models includes a convolution network from the pre-trained Resnet18 model, which extracts informative features from a high dimensional LISA traffic sign image dataset. The output from the classical processor is processed further through the quantum layer, which is composed of various quantum gates and provides support to various quantum mechanical features like entanglement and superposition. We have tested multiple combinations of quantum circuits to provide better classification accuracy with decreasing training data and found better resiliency for our hybrid classical-quantum deep learning model during attacks compared to the classical-only machine learning models.
Cell image analysis is crucial in Alzheimer's research to detect the presence of A$\beta$ protein inhibiting cell function. Deep learning speeds up the process by making only low-level data sufficient for fruitful inspection. We first found Unet is most suitable in augmented microscopy by comparing performance in multi-class semantics segmentation. We develop the augmented microscopy method to capture nuclei in a brightfield image and the transformer using Unet model to convert an input image into a sequence of topological information. The performance regarding Intersection-over-Union is consistent concerning the choice of image preprocessing and ground-truth generation. Training model with data of a specific cell type demonstrates transfer learning applies to some extent. The topological transformer aims to extract persistence silhouettes or landscape signatures containing geometric information of a given image of cells. This feature extraction facilitates studying an image as a collection of one-dimensional data, substantially reducing computational costs. Using the transformer, we attempt grouping cell images by their cell type relying solely on topological features. Performances of the transformers followed by SVM, XGBoost, LGBM, and simple convolutional neural network classifiers are inferior to the conventional image classification. However, since this research initiates a new perspective in biomedical research by combining deep learning and topology for image analysis, we speculate follow-up investigation will reinforce our genuine regime.
In real-life applications, machine learning models often face scenarios where there is a change in data distribution between training and test domains. When the aim is to make predictions on distributions different from those seen at training, we incur in a domain generalization problem. Methods to address this issue learn a model using data from multiple source domains, and then apply this model to the unseen target domain. Our hypothesis is that when training with multiple domains, conflicting gradients within each mini-batch contain information specific to the individual domains which is irrelevant to the others, including the test domain. If left untouched, such disagreement may degrade generalization performance. In this work, we characterize the conflicting gradients emerging in domain shift scenarios and devise novel gradient agreement strategies based on gradient surgery to alleviate their effect. We validate our approach in image classification tasks with three multi-domain datasets, showing the value of the proposed agreement strategy in enhancing the generalization capability of deep learning models in domain shift scenarios.
Sparse coding has been incorporated in models of the visual cortex for its computational advantages and connection to biology. But how the level of sparsity contributes to performance on visual tasks is not well understood. In this work, sparse coding has been integrated into an existing hierarchical V2 model (Hosoya and Hyv\"arinen, 2015), but replacing its independent component analysis (ICA) with an explicit sparse coding in which the degree of sparsity can be controlled. After training, the sparse coding basis functions with a higher degree of sparsity resembled qualitatively different structures, such as curves and corners. The contributions of the models were assessed with image classification tasks, specifically tasks associated with mid-level vision including figure-ground classification, texture classification, and angle prediction between two line stimuli. In addition, the models were assessed in comparison to a texture sensitivity measure that has been reported in V2 (Freeman et al., 2013), and a deleted-region inference task. The results from the experiments show that while sparse coding performed worse than ICA at classifying images, only sparse coding was able to better match the texture sensitivity level of V2 and infer deleted image regions, both by increasing the degree of sparsity in sparse coding. Higher degrees of sparsity allowed for inference over larger deleted image regions. The mechanism that allows for this inference capability in sparse coding is described here.
Vision transformers (ViTs) have recently received explosive popularity, but the huge computational cost is still a severe issue. Since the computation complexity of ViT is quadratic with respect to the input sequence length, a mainstream paradigm for computation reduction is to reduce the number of tokens. Existing designs include structured spatial compression that uses a progressive shrinking pyramid to reduce the computations of large feature maps, and unstructured token pruning that dynamically drops redundant tokens. However, the limitation of existing token pruning lies in two folds: 1) the incomplete spatial structure caused by pruning is not compatible with structured spatial compression that is commonly used in modern deep-narrow transformers; 2) it usually requires a time-consuming pre-training procedure. To tackle the limitations and expand the applicable scenario of token pruning, we present Evo-ViT, a self-motivated slow-fast token evolution approach for vision transformers. Specifically, we conduct unstructured instance-wise token selection by taking advantage of the simple and effective global class attention that is native to vision transformers. Then, we propose to update the selected informative tokens and uninformative tokens with different computation paths, namely, slow-fast updating. Since slow-fast updating mechanism maintains the spatial structure and information flow, Evo-ViT can accelerate vanilla transformers of both flat and deep-narrow structures from the very beginning of the training process. Experimental results demonstrate that our method significantly reduces the computational cost of vision transformers while maintaining comparable performance on image classification.
The importance of automated and objective monitoring of dietary behavior is becoming increasingly accepted. The advancements in sensor technology along with recent achievements in machine-learning--based signal-processing algorithms have enabled the development of dietary monitoring solutions that yield highly accurate results. A common bottleneck for developing and training machine learning algorithms is obtaining labeled data for training supervised algorithms, and in particular ground truth annotations. Manual ground truth annotation is laborious, cumbersome, can sometimes introduce errors, and is sometimes impossible in free-living data collection. As a result, there is a need to decrease the labeled data required for training. Additionally, unlabeled data, gathered in-the-wild from existing wearables (such as Bluetooth earbuds) can be used to train and fine-tune eating-detection models. In this work, we focus on training a feature extractor for audio signals captured by an in-ear microphone for the task of eating detection in a self-supervised way. We base our approach on the SimCLR method for image classification, proposed by Chen et al. from the domain of computer vision. Results are promising as our self-supervised method achieves similar results to supervised training alternatives, and its overall effectiveness is comparable to current state-of-the-art methods. Code is available at https://github.com/mug-auth/ssl-chewing .
Adversarial training based on the maximum classifier discrepancy between two classifier structures has achieved great success in unsupervised domain adaptation tasks for image classification. The approach adopts the structure of two classifiers, though simple and intuitive, the learned classification boundary may not well represent the data property in the new domain. In this paper, we propose to extend the structure to multiple classifiers to further boost its performance. To this end, we develop a very straightforward approach to adding more classifiers. We employ the principle that the classifiers are different from each other to construct a discrepancy loss function for multiple classifiers. The proposed construction method of loss function makes it possible to add any number of classifiers to the original framework. The proposed approach is validated through extensive experimental evaluations. We demonstrate that, on average, adopting the structure of three classifiers normally yields the best performance as a trade-off between accuracy and efficiency. With minimum extra computational costs, the proposed approach can significantly improve the performance of the original algorithm. The source code of the proposed approach can be downloaded from \url{https://github.com/rucv/MMCD\_DA}.
Deep learning based image classification models are shown vulnerable to adversarial attacks by injecting deliberately crafted noises to clean images. To defend against adversarial attacks in a training-free and attack-agnostic manner, this work proposes a novel and effective reconstruction-based defense framework by delving into deep image prior (DIP). Fundamentally different from existing reconstruction-based defenses, the proposed method analyzes and explicitly incorporates the model decision process into our defense. Given an adversarial image, firstly we map its reconstructed images during DIP optimization to the model decision space, where cross-boundary images can be detected and on-boundary images can be further localized. Then, adversarial noise is purified by perturbing on-boundary images along the reverse direction to the adversarial image. Finally, on-manifold images are stitched to construct an image that can be correctly predicted by the victim classifier. Extensive experiments demonstrate that the proposed method outperforms existing state-of-the-art reconstruction-based methods both in defending white-box attacks and defense-aware attacks. Moreover, the proposed method can maintain a high visual quality during adversarial image reconstruction.
Transformers have made great progress in dealing with computer vision tasks. However, existing vision transformers do not yet possess the ability of building the interactions among features of different scales, which is perceptually important to visual inputs. The reasons are two-fold: (1) Input embeddings of each layer are equal-scale, so no cross-scale feature can be extracted; (2) to lower the computational cost, some vision transformers merge adjacent embeddings inside the self-attention module, thus sacrificing small-scale (fine-grained) features of the embeddings and also disabling the cross-scale interactions. To this end, we propose Cross-scale Embedding Layer (CEL) and Long Short Distance Attention (LSDA). On the one hand, CEL blends each embedding with multiple patches of different scales, providing the self-attention module itself with cross-scale features. On the other hand, LSDA splits the self-attention module into a short-distance one and a long-distance counterpart, which not only reduces the computational burden but also keeps both small-scale and large-scale features in the embeddings. Through the above two designs, we achieve cross-scale attention. Besides, we put forward a dynamic position bias for vision transformers to make the popular relative position bias apply to variable-sized images. Hinging on the cross-scale attention module, we construct a versatile vision architecture, dubbed CrossFormer, which accommodates variable-sized inputs. Extensive experiments show that CrossFormer outperforms the other vision transformers on image classification, object detection, instance segmentation, and semantic segmentation tasks. The code has been released: https://github.com/cheerss/CrossFormer.
Semi-supervised image classification has shown substantial progress in learning from limited labeled data, but recent advances remain largely untested for clinical applications. Motivated by the urgent need to improve timely diagnosis of life-threatening heart conditions, especially aortic stenosis, we develop a benchmark dataset to assess semi-supervised approaches to two tasks relevant to cardiac ultrasound (echocardiogram) interpretation: view classification and disease severity classification. We find that a state-of-the-art method called MixMatch achieves promising gains in heldout accuracy on both tasks, learning from a large volume of truly unlabeled images as well as a labeled set collected at great expense to achieve better performance than is possible with the labeled set alone. We further pursue patient-level diagnosis prediction, which requires aggregating across hundreds of images of diverse view types, most of which are irrelevant, to make a coherent prediction. The best patient-level performance is achieved by new methods that prioritize diagnosis predictions from images that are predicted to be clinically-relevant views and transfer knowledge from the view task to the diagnosis task. We hope our released Tufts Medical Echocardiogram Dataset and evaluation framework inspire further improvements in multi-task semi-supervised learning for clinical applications.
In medical imaging, there are clinically relevant segmentation tasks where the output mask is a projection to a subset of input image dimensions. In this work, we propose a novel convolutional neural network architecture that can effectively learn to produce a lower-dimensional segmentation mask than the input image. The network restores encoded representation only in a subset of input spatial dimensions and keeps the representation unchanged in the others. The newly proposed projective skip-connections allow linking the encoder and decoder in a UNet-like structure. We evaluated the proposed method on two clinically relevant tasks in retinal Optical Coherence Tomography (OCT): geographic atrophy and retinal blood vessel segmentation. The proposed method outperformed the current state-of-the-art approaches on all the OCT datasets used, consisting of 3D volumes and corresponding 2D en-face masks. The proposed architecture fills the methodological gap between image classification and ND image segmentation.