Linear synthesis model based dictionary learning framework has achieved remarkable performances in image classification in the last decade. Behaved as a generative feature model, it however suffers from some intrinsic deficiencies. In this paper, we propose a novel parametric nonlinear analysis cosparse model (NACM) with which a unique feature vector will be much more efficiently extracted. Additionally, we derive a deep insight to demonstrate that NACM is capable of simultaneously learning the task adapted feature transformation and regularization to encode our preferences, domain prior knowledge and task oriented supervised information into the features. The proposed NACM is devoted to the classification task as a discriminative feature model and yield a novel discriminative nonlinear analysis operator learning framework (DNAOL). The theoretical analysis and experimental performances clearly demonstrate that DNAOL will not only achieve the better or at least competitive classification accuracies than the state-of-the-art algorithms but it can also dramatically reduce the time complexities in both training and testing phases.
Deriving from the gradient vector of a generative model of local features, Fisher vector coding (FVC) has been identified as an effective coding method for image classification. Most, if not all, FVC implementations employ the Gaussian mixture model (GMM) to depict the generation process of local features. However, the representative power of the GMM could be limited because it essentially assumes that local features can be characterized by a fixed number of feature prototypes and the number of prototypes is usually small in FVC. To handle this limitation, in this paper we break the convention which assumes that a local feature is drawn from one of few Gaussian distributions. Instead, we adopt a compositional mechanism which assumes that a local feature is drawn from a Gaussian distribution whose mean vector is composed as the linear combination of multiple key components and the combination weight is a latent random variable. In this way, we can greatly enhance the representative power of the generative model of FVC. To implement our idea, we designed two particular generative models with such a compositional mechanism.
Learning to learn plays a pivotal role in meta-learning (MTL) to obtain an optimal learning model. In this paper, we investigate mage recognition for unseen categories of a given dataset with limited training information. We deploy a zero-shot learning (ZSL) algorithm to achieve this goal. We also explore the effect of parameter tuning on performance of semantic auto-encoder (SAE). We further address the parameter tuning problem for meta-learning, especially focusing on zero-shot learning. By combining different embedded parameters, we improved the accuracy of tuned-SAE. Advantages and disadvantages of parameter tuning and its application in image classification are also explored.
This study aims to improve the control performance and generalization capability of end-to-end autonomous driving with scene understanding leveraging deep learning and multimodal sensor fusion technology. The designed end-to-end deep neural network takes the visual image and associated depth information as inputs in an early fusion level and outputs the pixel-wise semantic segmentation as scene understanding and vehicle control commands concurrently. The end-to-end deep learning-based autonomous driving model is tested in high-fidelity simulated urban driving conditions and compared with the benchmark of CoRL2017 and NoCrash. The testing results show that the proposed approach is of better performance and generalization ability, achieving a 100% success rate in static navigation tasks in both training and unobserved situations, as well as better success rates in other tasks than other existing models. A further ablation study shows that the model with the removal of multimodal sensor fusion or scene understanding pales in the new environment because of the false perception. The results verify that the performance of our model is improved by the synergy of multimodal sensor fusion with scene understanding subtask, demonstrating the feasibility and effectiveness of the developed deep neural network with multimodal sensor fusion.
In this paper we propose a novel supervised image classification method that overcomes dataset bias and scarcity of training data using privileged information in the form of keypoints annotations. Our main motivation is recognition of animal species for ecological applications like biodiversity modelling, which can be challenging due to long-tailed species distributions due to rare species, and strong dataset biases in repetitive scenes such as in camera traps. To counteract these challenges, we propose a weakly-supervised visual attention mechanism that has access to keypoints highlighting the most important object parts. This privileged information, implemented via a novel privileged pooling operation, is only accessible during training and helps the model to focus on the regions that are most discriminative. We show that the proposed approach uses more efficiently small training datasets, generalizes better and outperforms competing methods in challenging training conditions.
Recent advances in video manipulation techniques have made the generation of fake videos more accessible than ever before. Manipulated videos can fuel disinformation and reduce trust in media. Therefore detection of fake videos has garnered immense interest in academia and industry. Recently developed Deepfake detection methods rely on deep neural networks (DNNs) to distinguish AI-generated fake videos from real videos. In this work, we demonstrate that it is possible to bypass such detectors by adversarially modifying fake videos synthesized using existing Deepfake generation methods. We further demonstrate that our adversarial perturbations are robust to image and video compression codecs, making them a real-world threat. We present pipelines in both white-box and black-box attack scenarios that can fool DNN based Deepfake detectors into classifying fake videos as real.
Label hierarchies widely exist in many vision-related problems, ranging from explicit label hierarchies existed in image classification to latent label hierarchies existed in semantic segmentation. Nevertheless, state-of-the-art methods often deploy cross-entropy loss that implicitly assumes class labels to be exclusive and thus independence from each other. Motivated by the fact that classes from the same parental category usually share certain similarity, we design a new training diagram called Hierarchical Complement Objective Training (HCOT) that leverages the information from label hierarchy. HCOT maximizes the probability of the ground truth class, and at the same time, neutralizes the probabilities of rest of the classes in a hierarchical fashion, making the model take advantage of the label hierarchy explicitly. The proposed HCOT is evaluated on both image classification and semantic segmentation tasks. Experimental results confirm that HCOT outperforms state-of-the-art models in CIFAR-100, ImageNet-2012, and PASCAL-Context. The study further demonstrates that HCOT can be applied on tasks with latent label hierarchies, which is a common characteristic in many machine learning tasks.
Few-shot learning aims to classify unseen classes with a few training examples. While recent works have shown that standard mini-batch training with a carefully designed training strategy can improve generalization ability for unseen classes, well-known problems in deep networks such as memorizing training statistics have been less explored for few-shot learning. To tackle this issue, we propose self-augmentation that consolidates regional dropout and self-distillation. Specifically, we exploit a data augmentation technique called regional dropout, in which a patch of an image is substituted into other values. Then, we employ a backbone network that has auxiliary branches with its own classifier to enforce knowledge sharing. Lastly, we present a fine-tuning method to further exploit a few training examples for unseen classes. Experimental results show that the proposed method outperforms the state-of-the-art methods for prevalent few-shot benchmarks and improves the generalization ability.
In this paper, we demonstrate a physical adversarial patch attack against object detectors, notably the YOLOv3 detector. Unlike previous work on physical object detection attacks, which required the patch to overlap with the objects being misclassified or avoiding detection, we show that a properly designed patch can suppress virtually all the detected objects in the image. That is, we can place the patch anywhere in the image, causing all existing objects in the image to be missed entirely by the detector, even those far away from the patch itself. This in turn opens up new lines of physical attacks against object detection systems, which require no modification of the objects in a scene. A demo of the system can be found at https://youtu.be/WXnQjbZ1e7Y.
Remote sensing image classification can be performed in many different ways to extract meaningful features. One common approach is to perform edge detection. A second approach is to try and detect whole shapes, given the fact that these shapes usually tend to have distinctive properties such as object foreground or background. To get optimal results, these two approaches can be combined. This paper adopts a combinatorial optimization method to adaptively select threshold based features to improve remote sensing image. Feature selection is an important combinatorial optimization problem in the remote sensing image classification. The feature selection method has to achieve three characteristics: first the performance issues by facilitating data collection and reducing storage space and classification time, second to perform semantics analysis helping to understand the problem, and third to improve prediction accuracy by avoiding the curse of dimensionality. The goal of this thresholding an image is to classify pixels as either dark or light and evaluation of classification results. Interactive adaptive thresholding is a form of thresholding that takes into account spatial variations in illumination of remote sensing image. We present a technique for remote sensing based adaptive thresholding using the interactive satellite image of the input. However, our solution is more robust to illumination changes in the remote sensing image. Additionally, our method is simple and easy to implement but it is effective algorithm to classify the image pixels. This technique is suitable for preprocessing the remote sensing image classification, making it a valuable tool for interactive remote based applications such as augmented reality of the classification procedure.