Recent works have advanced the performance of self-supervised representation learning by a large margin. The core among these methods is intra-image invariance learning. Two different transformations of one image instance are considered as a positive sample pair, where various tasks are designed to learn invariant representations by comparing the pair. Analogically, for video data, representations of frames from the same video are trained to be closer than frames from other videos, i.e. intra-video invariance. However, cross-video relation has barely been explored for visual representation learning. Unlike intra-video invariance, ground-truth labels of cross-video relation is usually unavailable without human labors. In this paper, we propose a novel contrastive learning method which explores the cross-video relation by using cycle-consistency for general image representation learning. This allows to collect positive sample pairs across different video instances, which we hypothesize will lead to higher-level semantics. We validate our method by transferring our image representation to multiple downstream tasks including visual object tracking, image classification, and action recognition. We show significant improvement over state-of-the-art contrastive learning methods. Project page is available at https://happywu.github.io/cycle_contrast_video.
Convolutional Neural Networks (ConvNets) are trained offline using the few available data and may therefore suffer from substantial accuracy loss when ported on the field, where unseen input patterns received under unpredictable external conditions can mislead the model. Test-Time Augmentation (TTA) techniques aim to alleviate such common side effect at inference-time, first running multiple feed-forward passes on a set of altered versions of the same input sample, and then computing the main outcome through a consensus of the aggregated predictions. Unfortunately, the implementation of TTA on embedded CPUs introduces latency penalties that limit its adoption on edge applications. To tackle this issue, we propose AdapTTA, an adaptive implementation of TTA that controls the number of feed-forward passes dynamically, depending on the complexity of the input. Experimental results on state-of-the-art ConvNets for image classification deployed on a commercial ARM Cortex-A CPU demonstrate AdapTTA reaches remarkable latency savings, from 1.49X to 2.21X, and hence a higher frame rate compared to static TTA, still preserving the same accuracy gain.
Multi-domain image-to-image translation with conditional Generative Adversarial Networks (GANs) can generate highly photo realistic images with desired target classes, yet these synthetic images have not always been helpful to improve downstream supervised tasks such as image classification. Improving downstream tasks with synthetic examples requires generating images with high fidelity to the unknown conditional distribution of the target class, which many labeled conditional GANs attempt to achieve by adding soft-max cross-entropy loss based auxiliary classifier in the discriminator. As recent studies suggest that the soft-max loss in Euclidean space of deep feature does not leverage their intrinsic angular distribution, we propose to replace this loss in auxiliary classifier with an additive angular margin (AAM) loss that takes benefit of the intrinsic angular distribution, and promotes intra-class compactness and inter-class separation to help generator synthesize high fidelity images. We validate our method on RaFD and CIFAR-100, two challenging face expression and natural image classification data set. Our method outperforms state-of-the-art methods in several different evaluation criteria including recently proposed GAN-train and GAN-test metrics designed to assess the impact of synthetic data on downstream classification task, assessing the usefulness in data augmentation for supervised tasks with prediction accuracy score and average confidence score, and the well known FID metric.
Collecting a large number of reliable training images annotated by multiple land-cover class labels in the framework of multi-label classification is time-consuming and costly in remote sensing (RS). To address this problem, publicly available thematic products are often used for annotating RS images with zero-labeling-cost. However, such an approach may result in constructing a training set with noisy multi-labels, distorting the learning process. To address this problem, we propose a Consensual Collaborative Multi-Label Learning (CCML) method. The proposed CCML identifies, ranks and corrects training images with noisy multi-labels through four main modules: 1) discrepancy module; 2) group lasso module; 3) flipping module; and 4) swap module. The discrepancy module ensures that the two networks learn diverse features, while obtaining the same predictions. The group lasso module detects the potentially noisy labels by estimating the label uncertainty based on the aggregation of two collaborative networks. The flipping module corrects the identified noisy labels, whereas the swap module exchanges the ranking information between the two networks. The experimental results confirm the success of the proposed CCML under high (synthetically added) multi-label noise rates. The code of the proposed method is publicly available at https://noisy-labels-in-rs.org
Image segmentation is often ambiguous at the level of individual image patches and requires contextual information to reach label consensus. In this paper we introduce Segmenter, a transformer model for semantic segmentation. In contrast to convolution-based methods, our approach allows to model global context already at the first layer and throughout the network. We build on the recent Vision Transformer (ViT) and extend it to semantic segmentation. To do so, we rely on the output embeddings corresponding to image patches and obtain class labels from these embeddings with a point-wise linear decoder or a mask transformer decoder. We leverage models pre-trained for image classification and show that we can fine-tune them on moderate sized datasets available for semantic segmentation. The linear decoder allows to obtain excellent results already, but the performance can be further improved by a mask transformer generating class masks. We conduct an extensive ablation study to show the impact of the different parameters, in particular the performance is better for large models and small patch sizes. Segmenter attains excellent results for semantic segmentation. It outperforms the state of the art on both ADE20K and Pascal Context datasets and is competitive on Cityscapes.
To classify images based on their content is one of the most studied topics in the field of computer vision. Nowadays, this problem can be addressed using modern techniques such as Convolutional Neural Networks (CNN), but over the years different classical methods have been developed. In this report, we implement an image classifier using both classic computer vision and deep learning techniques. Specifically, we study the performance of a Bag of Visual Words classifier using Support Vector Machines, a Multilayer Perceptron, an existing architecture named InceptionV3 and our own CNN, TinyNet, designed from scratch. We evaluate each of the cases in terms of accuracy and loss, and we obtain results that vary between 0.6 and 0.96 depending on the model and configuration used.
Convolutional neural networks (CNNs) have been applied to learn spatial features for high-resolution (HR) synthetic aperture radar (SAR) image classification. However, there has been little work on integrating the unique statistical distributions of SAR images which can reveal physical properties of terrain objects, into CNNs in a supervised feature learning framework. To address this problem, a novel end-to-end supervised classification method is proposed for HR SAR images by considering both spatial context and statistical features. First, to extract more effective spatial features from SAR images, a new deep spatial context encoder network (DSCEN) is proposed, which is a lightweight structure and can be effectively trained with a small number of samples. Meanwhile, to enhance the diversity of statistics, the nonstationary joint statistical model (NS-JSM) is adopted to form the global statistical features. Specifically, SAR images are transformed into the Gabor wavelet domain and the produced multi-subbands magnitudes and phases are modeled by the log-normal and uniform distribution. The covariance matrix is further utilized to capture the inter-scale and intra-scale nonstationary correlation between the statistical subbands and make the joint statistical features more compact and distinguishable. Considering complementary advantages, a feature fusion network (Fusion-Net) base on group compression and smooth normalization is constructed to embed the statistical features into the spatial features and optimize the fusion feature representation. As a result, our model can learn the discriminative features and improve the final classification performance. Experiments on four HR SAR images validate the superiority of the proposed method over other related algorithms.
Convolutional Neural Networks (CNNs) specialize in feature extraction rather than function mapping. In doing so they form complex internal hierarchical feature representations, the complexity of which gradually increases with a corresponding increment in neural network depth. In this paper, we examine the feature extraction capabilities of CNNs using Maximum Entropy (ME) and Signal-to-Noise Ratio (SNR) to validate the idea that, CNN models should be tailored for a given task and complexity of the input data. SNR and ME measures are used as they can accurately determine in the input dataset, the relative amount of signal information to the random noise and the maximum amount of information respectively. We use two well known benchmarking datasets, MNIST and CIFAR-10 to examine the information extraction and abstraction capabilities of CNNs. Through our experiments, we examine convolutional feature extraction and abstraction capabilities in CNNs and show that the classification accuracy or performance of CNNs is greatly dependent on the amount, complexity and quality of the signal information present in the input data. Furthermore, we show the effect of information overflow and underflow on CNN classification accuracies. Our hypothesis is that the feature extraction and abstraction capabilities of convolutional layers are limited and therefore, CNN models should be tailored to the input data by using appropriately sized CNNs based on the SNR and ME measures of the input dataset.
The connection between training deep neural networks (DNNs) and optimal control theory (OCT) has attracted considerable attention as a principled tool of algorithmic design. Despite few attempts being made, they have been limited to architectures where the layer propagation resembles a Markovian dynamical system. This casts doubts on their flexibility to modern networks that heavily rely on non-Markovian dependencies between layers (e.g. skip connections in residual networks). In this work, we propose a novel dynamic game perspective by viewing each layer as a player in a dynamic game characterized by the DNN itself. Through this lens, different classes of optimizers can be seen as matching different types of Nash equilibria, depending on the implicit information structure of each (p)layer. The resulting method, called Dynamic Game Theoretic Neural Optimizer (DGNOpt), not only generalizes OCT-inspired optimizers to richer network class; it also motivates a new training principle by solving a multi-player cooperative game. DGNOpt shows convergence improvements over existing methods on image classification datasets with residual and inception networks. Our work marries strengths from both OCT and game theory, paving ways to new algorithmic opportunities from robust optimal control and bandit-based optimization.
We propose a novel clustering mechanism based on an incompatibility property between subsets of data that emerges during model training. This mechanism partitions the dataset into subsets that generalize only to themselves, i.e., training on one subset does not improve performance on the other subsets. Leveraging the interaction between the dataset and the training process, our clustering mechanism partitions datasets into clusters that are defined by--and therefore meaningful to--the objective of the training process. We apply our clustering mechanism to defend against data poisoning attacks, in which the attacker injects malicious poisoned data into the training dataset to affect the trained model's output. Our evaluation focuses on backdoor attacks against deep neural networks trained to perform image classification using the GTSRB and CIFAR-10 datasets. Our results show that (1) these attacks produce poisoned datasets in which the poisoned and clean data are incompatible and (2) our technique successfully identifies (and removes) the poisoned data. In an end-to-end evaluation, our defense reduces the attack success rate to below 1% on 134 out of 165 scenarios, with only a 2% drop in clean accuracy on CIFAR-10 and a negligible drop in clean accuracy on GTSRB.
We propose a learning framework based on stochastic Bregman iterations, also known as mirror descent, to train sparse neural networks with an inverse scale space approach. We derive a baseline algorithm called LinBreg, an accelerated version using momentum, and AdaBreg, which is a Bregmanized generalization of the Adam algorithm. In contrast to established methods for sparse training the proposed family of algorithms constitutes a regrowth strategy for neural networks that is solely optimization-based without additional heuristics. Our Bregman learning framework starts the training with very few initial parameters, successively adding only significant ones to obtain a sparse and expressive network. The proposed approach is extremely easy and efficient, yet supported by the rich mathematical theory of inverse scale space methods. We derive a statistically profound sparse parameter initialization strategy and provide a rigorous stochastic convergence analysis of the loss decay and additional convergence proofs in the convex regime. Using only 3.4% of the parameters of ResNet-18 we achieve 90.2% test accuracy on CIFAR-10, compared to 93.6% using the dense network. Our algorithm also unveils an autoencoder architecture for a denoising task. The proposed framework also has a huge potential for integrating sparse backpropagation and resource-friendly training.
This paper presents AppealNet, a novel edge/cloud collaborative architecture that runs deep learning (DL) tasks more efficiently than state-of-the-art solutions. For a given input, AppealNet accurately predicts on-the-fly whether it can be successfully processed by the DL model deployed on the resource-constrained edge device, and if not, appeals to the more powerful DL model deployed at the cloud. This is achieved by employing a two-head neural network architecture that explicitly takes inference difficulty into consideration and optimizes the tradeoff between accuracy and computation/communication cost of the edge/cloud collaborative architecture. Experimental results on several image classification datasets show up to more than 40% energy savings compared to existing techniques without sacrificing accuracy.
Few-shot learning deals with the fundamental and challenging problem of learning from a few annotated samples, while being able to generalize well on new tasks. The crux of few-shot learning is to extract prior knowledge from related tasks to enable fast adaptation to a new task with a limited amount of data. In this paper, we propose meta-learning kernels with random Fourier features for few-shot learning, we call MetaKernel. Specifically, we propose learning variational random features in a data-driven manner to obtain task-specific kernels by leveraging the shared knowledge provided by related tasks in a meta-learning setting. We treat the random feature basis as the latent variable, which is estimated by variational inference. The shared knowledge from related tasks is incorporated into a context inference of the posterior, which we achieve via a long-short term memory module. To establish more expressive kernels, we deploy conditional normalizing flows based on coupling layers to achieve a richer posterior distribution over random Fourier bases. The resultant kernels are more informative and discriminative, which further improves the few-shot learning. To evaluate our method, we conduct extensive experiments on both few-shot image classification and regression tasks. A thorough ablation study demonstrates that the effectiveness of each introduced component in our method. The benchmark results on fourteen datasets demonstrate MetaKernel consistently delivers at least comparable and often better performance than state-of-the-art alternatives.
Prior works have found it beneficial to combine provably noise-robust loss functions e.g., mean absolute error (MAE) with standard categorical loss function e.g. cross entropy (CE) to improve their learnability. Here, we propose to use Jensen-Shannon divergence as a noise-robust loss function and show that it interestingly interpolate between CE and MAE with a controllable mixing parameter. Furthermore, we make a crucial observation that CE exhibit lower consistency around noisy data points. Based on this observation, we adopt a generalized version of the Jensen-Shannon divergence for multiple distributions to encourage consistency around data points. Using this loss function, we show state-of-the-art results on both synthetic (CIFAR), and real-world (e.g., WebVision) noise with varying noise rates.
Recently, Convolutional Neural Networks (CNNs) have made a great performance for remote sensing image classification. Plant recognition using CNNs is one of the active deep learning research topics due to its added-value in different related fields, especially environmental conservation and natural areas preservation. Automatic recognition of plants in protected areas helps in the surveillance process of these zones and ensures the sustainability of their ecosystems. In this work, we propose an Enhanced Randomly Initialized Convolutional Neural Network (ERI-CNN) for the recognition of columnar cactus, which is an endemic plant that exists in the Tehuac\'an-Cuicatl\'an Valley in southeastern Mexico. We used a public dataset created by a group of researchers that consists of more than 20000 remote sensing images. The experimental results confirm the effectiveness of the proposed model compared to other models reported in the literature like InceptionV3 and the modified LeNet-5 CNN. Our ERI-CNN provides 98% of accuracy, 97% of precision, 97% of recall, 97.5% as f1-score, and 0.056 loss.
Vision systems, i.e., systems that allow to detect and track objects in images, have gained substantial importance over the past decades. They are used in quality assurance applications, e.g., for finding surface defects in products during manufacturing, surveillance, but also automated driving, requiring reliable behavior. Interestingly, there is only little work on quality assurance and especially testing of vision systems in general. In this paper, we contribute to the area of testing vision software, and present a framework for the automated generation of tests for systems based on vision and image recognition. The framework makes use of existing libraries allowing to modify original images and to obtain similarities between the original and modified images. We show how such a framework can be used for testing a particular industrial application on identifying defects on riblet surfaces and present preliminary results from the image classification domain.
Convolutional Neural Networks (CNNs) have achieved tremendous success in a number of learning tasks including image classification. Recent advanced models in CNNs, such as ResNets, mainly focus on the skip connection to avoid gradient vanishing. DenseNet designs suggest creating additional bypasses to transfer features as an alternative strategy in network design. In this paper, we design Attentive Feature Integration (AFI) modules, which are widely applicable to most recent network architectures, leading to new architectures named AFI-Nets. AFI-Nets explicitly model the correlations among different levels of features and selectively transfer features with a little overhead.AFI-ResNet-152 obtains a 1.24% relative improvement on the ImageNet dataset while decreases the FLOPs by about 10% and the number of parameters by about 9.2% compared to ResNet-152.
Previous robustness approaches for deep learning models such as data augmentation techniques via data transformation or adversarial training cannot capture real-world variations that preserve the semantics of the input, such as a change in lighting conditions. To bridge this gap, we present NaTra, an adversarial training scheme that is designed to improve the robustness of image classification algorithms. We target attributes of the input images that are independent of the class identification, and manipulate those attributes to mimic real-world natural transformations (NaTra) of the inputs, which are then used to augment the training dataset of the image classifier. Specifically, we apply \textit{Batch Inverse Encoding and Shifting} to map a batch of given images to corresponding disentangled latent codes of well-trained generative models. \textit{Latent Codes Expansion} is used to boost image reconstruction quality through the incorporation of extended feature maps. \textit{Unsupervised Attribute Directing and Manipulation} enables identification of the latent directions that correspond to specific attribute changes, and then produce interpretable manipulations of those attributes, thereby generating natural transformations to the input data. We demonstrate the efficacy of our scheme by utilizing the disentangled latent representations derived from well-trained GANs to mimic transformations of an image that are similar to real-world natural variations (such as lighting conditions or hairstyle), and train models to be invariant to these natural transformations. Extensive experiments show that our method improves generalization of classification models and increases its robustness to various real-world distortions
Convolutional neural networks lack shift equivariance due to the presence of downsampling layers. In image classification, adaptive polyphase downsampling (APS-D) was recently proposed to make CNNs perfectly shift invariant. However, in networks used for image reconstruction tasks, it can not by itself restore shift equivariance. We address this problem by proposing adaptive polyphase upsampling (APS-U), a non-linear extension of conventional upsampling, which allows CNNs with symmetric encoder-decoder architecture (for example U-Net) to exhibit perfect shift equivariance. With MRI and CT reconstruction experiments, we show that networks containing APS-D/U layers exhibit state of the art equivariance performance without sacrificing on image reconstruction quality. In addition, unlike prior methods like data augmentation and anti-aliasing, the gains in equivariance obtained from APS-D/U also extend to images outside the training distribution.
Cancer diseases constitute one of the most significant societal challenges. In this paper, we introduce a novel histopathological dataset for prostate cancer detection. The proposed dataset, consisting of over 2.6 million tissue patches extracted from 430 fully annotated scans, 4675 scans with assigned binary diagnoses, and 46 scans with diagnoses independently provided by a group of histopathologists can be found at https://github.com/michalkoziarski/DiagSet. Furthermore, we propose a machine learning framework for detection of cancerous tissue regions and prediction of scan-level diagnosis, utilizing thresholding to abstain from the decision in uncertain cases. The proposed approach, composed of ensembles of deep neural networks operating on the histopathological scans at different scales, achieves 94.6% accuracy in patch-level recognition and is compared in a scan-level diagnosis with 9 human histopathologists showing high statistical agreement.