Few-shot object detection (FSOD) aims to classify and detect few images of novel categories. Existing meta-learning methods insufficiently exploit features between support and query images owing to structural limitations. We propose a hierarchical attention network with sequentially large receptive fields to fully exploit the query and support images. In addition, meta-learning does not distinguish the categories well because it determines whether the support and query images match. In other words, metric-based learning for classification is ineffective because it does not work directly. Thus, we propose a contrastive learning method called meta-contrastive learning, which directly helps achieve the purpose of the meta-learning strategy. Finally, we establish a new state-of-the-art network, by realizing significant margins. Our method brings 2.3, 1.0, 1.3, 3.4 and 2.4% AP improvements for 1-30 shots object detection on COCO dataset. Our code is available at: https://github.com/infinity7428/hANMCL
Automatic detection of traffic accidents is an important emerging topic in traffic monitoring systems. Nowadays many urban intersections are equipped with surveillance cameras connected to traffic management systems. Therefore, computer vision techniques can be viable tools for automatic accident detection. This paper presents a new efficient framework for accident detection at intersections for traffic surveillance applications. The proposed framework consists of three hierarchical steps, including efficient and accurate object detection based on the state-of-the-art YOLOv4 method, object tracking based on Kalman filter coupled with the Hungarian algorithm for association, and accident detection by trajectory conflict analysis. A new cost function is applied for object association to accommodate for occlusion, overlapping objects, and shape changes in the object tracking step. The object trajectories are analyzed in terms of velocity, angle, and distance in order to detect different types of trajectory conflicts including vehicle-to-vehicle, vehicle-to-pedestrian, and vehicle-to-bicycle. Experimental results using real traffic video data show the feasibility of the proposed method in real-time applications of traffic surveillance. In particular, trajectory conflicts, including near-accidents and accidents occurring at urban intersections are detected with a low false alarm rate and a high detection rate. The robustness of the proposed framework is evaluated using video sequences collected from YouTube with diverse illumination conditions. The dataset is publicly available at: http://github.com/hadi-ghnd/AccidentDetection.
\noindent Traditional L_p norm-restricted image attack algorithms suffer from poor transferability to black box scenarios and poor robustness to defense algorithms. Recent CNN generator-based attack approaches can synthesize unrestricted and semantically meaningful entities to the image, which is shown to be transferable and robust. However, such methods attack images by either synthesizing local adversarial entities, which are only suitable for attacking specific contents or performing global attacks, which are only applicable to a specific image scale. In this paper, we propose a novel Patch Quilting Generative Adversarial Networks (PQ-GAN) to learn the first scale-free CNN generator that can be applied to attack images with arbitrary scales for various computer vision tasks. The principal investigation on transferability of the generated adversarial examples, robustness to defense frameworks, and visual quality assessment show that the proposed PQG-based attack framework outperforms the other nine state-of-the-art adversarial attack approaches when attacking the neural networks trained on two standard evaluation datasets (i.e., ImageNet and CityScapes).
Contrastive learning is commonly applied to self-supervised learning, and has been shown to outperform traditional approaches such as the triplet loss and N-pair loss. However, the requirement of large batch sizes and memory banks has made it difficult and slow to train. Recently, Supervised Contrasative approaches have been developed to overcome these problems. They focus more on learning a good representation for each class individually, or between a cluster of classes. In this work we attempt to rank classes based on similarity using a user-defined ranking, to learn an efficient representation between all classes. We observe how incorporating human bias into the learning process could improve learning representations in the parameter space. We show that our results are comparable to Supervised Contrastive Learning for image classification and object detection, and discuss it's shortcomings in OOD Detection
Unlike the conventional Knowledge Distillation (KD), Self-KD allows a network to learn knowledge from itself without any guidance from extra networks. This paper proposes to perform Self-KD from image Mixture (MixSKD), which integrates these two techniques into a unified framework. MixSKD mutually distills feature maps and probability distributions between the random pair of original images and their mixup images in a meaningful way. Therefore, it guides the network to learn cross-image knowledge by modelling supervisory signals from mixup images. Moreover, we construct a self-teacher network by aggregating multi-stage feature maps for providing soft labels to supervise the backbone classifier, further improving the efficacy of self-boosting. Experiments on image classification and transfer learning to object detection and semantic segmentation demonstrate that MixSKD outperforms other state-of-the-art Self-KD and data augmentation methods. The code is available at https://github.com/winycg/Self-KD-Lib.
To boost the performance, deep neural networks require deeper or wider network structures that involve massive computational and memory costs. To alleviate this issue, the self-knowledge distillation method regularizes the model by distilling the internal knowledge of the model itself. Conventional self-knowledge distillation methods require additional trainable parameters or are dependent on the data. In this paper, we propose a simple and effective self-knowledge distillation method using a dropout (SD-Dropout). SD-Dropout distills the posterior distributions of multiple models through a dropout sampling. Our method does not require any additional trainable modules, does not rely on data, and requires only simple operations. Furthermore, this simple method can be easily combined with various self-knowledge distillation approaches. We provide a theoretical and experimental analysis of the effect of forward and reverse KL-divergences in our work. Extensive experiments on various vision tasks, i.e., image classification, object detection, and distribution shift, demonstrate that the proposed method can effectively improve the generalization of a single network. Further experiments show that the proposed method also improves calibration performance, adversarial robustness, and out-of-distribution detection ability.
Capturing an event from multiple camera angles can give a viewer the most complete and interesting picture of that event. To be suitable for broadcasting, a human director needs to decide what to show at each point in time. This can become cumbersome with an increasing number of camera angles. The introduction of omnidirectional or wide-angle cameras has allowed for events to be captured more completely, making it even more difficult for the director to pick a good shot. In this paper, a system is presented that, given multiple ultra-high resolution video streams of an event, can generate a visually pleasing sequence of shots that manages to follow the relevant action of an event. Due to the algorithm being general purpose, it can be applied to most scenarios that feature humans. The proposed method allows for online processing when real-time broadcasting is required, as well as offline processing when the quality of the camera operation is the priority. Object detection is used to detect humans and other objects of interest in the input streams. Detected persons of interest, along with a set of rules based on cinematic conventions, are used to determine which video stream to show and what part of that stream is virtually framed. The user can provide a number of settings that determine how these rules are interpreted. The system is able to handle input from different wide-angle video streams by removing lens distortions. Using a user study it is shown, for a number of different scenarios, that the proposed automated director is able to capture an event with aesthetically pleasing video compositions and human-like shot switching behavior.
We recently proposed a new cluster operating system stack, DBOS, centered on a DBMS. DBOS enables unique support for ML applications by encapsulating ML code within stored procedures, centralizing ancillary ML data, providing security built into the underlying DBMS, co-locating ML code and data, and tracking data and workflow provenance. Here we demonstrate a subset of these benefits around two ML applications. We first show that image classification and object detection models using GPUs can be served as DBOS stored procedures with performance competitive to existing systems. We then present a 1D CNN trained to detect anomalies in HTTP requests on DBOS-backed web services, achieving SOTA results. We use this model to develop an interactive anomaly detection system and evaluate it through qualitative user feedback, demonstrating its usefulness as a proof of concept for future work to develop learned real-time security services on top of DBOS.
Continuous Speech Keyword Spotting (CSKWS) is a task to detect predefined keywords in a continuous speech. In this paper, we regard CSKWS as a one-dimensional object detection task and propose a novel anchor-free detector, named AF-KWS, to solve the problem. AF-KWS directly regresses the center locations and lengths of the keywords through a single-stage deep neural network. In particular, AF-KWS is tailored for this speech task as we introduce an auxiliary unknown class to exclude other words from non-speech or silent background. We have built two benchmark datasets named LibriTop-20 and continuous meeting analysis keywords (CMAK) dataset for CSKWS. Evaluations on these two datasets show that our proposed AF-KWS outperforms reference schemes by a large margin, and therefore provides a decent baseline for future research.
Object localization has been a crucial task in computer vision field. Methods of localizing objects in an image have been proposed based on the features of the attended pixels. Recently researchers have proposed methods to formulate object localization as a dynamic decision process, which can be solved by a reinforcement learning approach. In this project, we implement a novel active object localization algorithm based on deep reinforcement learning. We compare two different action settings for this MDP: a hierarchical method and a dynamic method. We further perform some ablation studies on the performance of the models by investigating different hyperparameters and various architecture changes.
Nondestructive testing (NDT) is widely applied to defect identification of turbine components during manufacturing and operation. Operational efficiency is key for gas turbine OEM (Original Equipment Manufacturers). Automating the inspection process as much as possible, while minimizing the uncertainties involved, is thus crucial. We propose a model based on RetinaNet to identify drilling defects in X-ray images of turbine blades. The application is challenging due to the large image resolutions in which defects are very small and hardly captured by the commonly used anchor sizes, and also due to the small size of the available dataset. As a matter of fact, all these issues are pretty common in the application of Deep Learning-based object detection models to industrial defect data. We overcome such issues using open source models, splitting the input images into tiles and scaling them up, applying heavy data augmentation, and optimizing the anchor size and aspect ratios with a differential evolution solver. We validate the model with $3$-fold cross-validation, showing a very high accuracy in identifying images with defects. We also define a set of best practices which can help other practitioners overcome similar challenges.
Salient Object Detection (SOD) is a popular and important topic aimed at precise detection and segmentation of the interesting regions in the images. We integrate the linguistic information into the vision-based U-Structure networks designed for salient object detection tasks. The experiments are based on the newly created DUTS Cross Modal (DUTS-CM) dataset, which contains both visual and linguistic labels. We propose a new module called efficient Cross-Modal Self-Attention (eCMSA) to combine visual and linguistic features and improve the performance of the original U-structure networks. Meanwhile, to reduce the heavy burden of labeling, we employ a semi-supervised learning method by training an image caption model based on the DUTS-CM dataset, which can automatically label other datasets like DUT-OMRON and HKU-IS. The comprehensive experiments show that the performance of SOD can be improved with the natural language input and is competitive compared with other SOD methods.
Images captured in real-world applications in remote sensing, image or video retrieval, and outdoor surveillance suffer degraded quality introduced by poor weather conditions. Conditions such as rain and mist, introduce artifacts that make visual analysis challenging and limit the performance of high-level computer vision methods. For time-critical applications where a rapid response is necessary, it becomes crucial to develop algorithms that automatically remove rain, without diminishing the quality of the image contents. This article aims to develop a novel quaternion multi-stage multiscale neural network with a self-attention module called QSAM-Net to remove rain streaks. The novelty of this algorithm is that it requires significantly fewer parameters by a factor of 3.98, over prior methods, while improving visual quality. This is demonstrated by the extensive evaluation and benchmarking on synthetic and real-world rainy images. This feature of QSAM-Net makes the network suitable for implementation on edge devices and applications requiring near real-time performance. The experiments demonstrate that by improving the visual quality of images. In addition, object detection accuracy and training speed are also improved.
Finding quantitative descriptors representing the microstructural features of a given material is an ongoing research area in the paradigm of Materials-by-Design. Historically, microstructural analysis mostly relies on qualitative descriptions. However, to build a robust and accurate process-structure-properties relationship, which is required for designing new advanced high-performance materials, the extraction of quantitative and meaningful statistical data from the microstructural analysis is a critical step. In recent years, computer vision (CV) methods, especially those which are centered around convolutional neural network (CNN) algorithms have shown promising results for this purpose. This review paper focuses on the state-of-the-art CNN-based techniques that have been applied to various multi-scale microstructural image analysis tasks, including classification, object detection, segmentation, feature extraction, and reconstruction. Additionally, we identified the main challenges with regard to the application of these methods to materials science research. Finally, we discussed some possible future directions of research in this area. In particular, we emphasized the application of transformer-based models and their capabilities to improve the microstructural analysis of materials.
Convolutional neural networks (CNNs) have made resounding success in many computer vision tasks such as image classification and object detection. However, their performance degrades rapidly on tougher tasks where images are of low resolution or objects are small. In this paper, we point out that this roots in a defective yet common design in existing CNN architectures, namely the use of strided convolution and/or pooling layers, which results in a loss of fine-grained information and learning of less effective feature representations. To this end, we propose a new CNN building block called SPD-Conv in place of each strided convolution layer and each pooling layer (thus eliminates them altogether). SPD-Conv is comprised of a space-to-depth (SPD) layer followed by a non-strided convolution (Conv) layer, and can be applied in most if not all CNN architectures. We explain this new design under two most representative computer vision tasks: object detection and image classification. We then create new CNN architectures by applying SPD-Conv to YOLOv5 and ResNet, and empirically show that our approach significantly outperforms state-of-the-art deep learning models, especially on tougher tasks with low-resolution images and small objects. We have open-sourced our code at https://github.com/LabSAINT/SPD-Conv.
Two-stage detectors have gained much popularity in 3D object detection. Most two-stage 3D detectors utilize grid points, voxel grids, or sampled keypoints for RoI feature extraction in the second stage. Such methods, however, are inefficient in handling unevenly distributed and sparse outdoor points. This paper solves this problem in three aspects. 1) Dynamic Point Aggregation. We propose the patch search to quickly search points in a local region for each 3D proposal. The dynamic farthest voxel sampling is then applied to evenly sample the points. Especially, the voxel size varies along the distance to accommodate the uneven distribution of points. 2) RoI-graph Pooling. We build local graphs on the sampled points to better model contextual information and mine point relations through iterative message passing. 3) Visual Features Augmentation. We introduce a simple yet effective fusion strategy to compensate for sparse LiDAR points with limited semantic cues. Based on these modules, we construct our Graph R-CNN as the second stage, which can be applied to existing one-stage detectors to consistently improve the detection performance. Extensive experiments show that Graph R-CNN outperforms the state-of-the-art 3D detection models by a large margin on both the KITTI and Waymo Open Dataset. And we rank first place on the KITTI BEV car detection leaderboard. Code will be available at \url{https://github.com/Nightmare-n/GraphRCNN}.
Robots working in unstructured environments must be capable of sensing and interpreting their surroundings. One of the main obstacles of deep-learning-based models in the field of robotics is the lack of domain-specific labeled data for different industrial applications. In this article, we propose a sim2real transfer learning method based on domain randomization for object detection with which labeled synthetic datasets of arbitrary size and object types can be automatically generated. Subsequently, a state-of-the-art convolutional neural network, YOLOv4, is trained to detect the different types of industrial objects. With the proposed domain randomization method, we could shrink the reality gap to a satisfactory level, achieving 86.32% and 97.38% mAP50 scores, respectively, in the case of zero-shot and one-shot transfers, on our manually annotated dataset containing 190 real images. Our solution fits for industrial use as the data generation process takes less than 0.5 s per image and the training lasts only around 12 h, on a GeForce RTX 2080 Ti GPU. Furthermore, it can reliably differentiate similar classes of objects by having access to only one real image for training. To our best knowledge, this is the only work thus far satisfying these constraints.
Drones equipped with cameras can significantly enhance human ability to perceive the world because of their remarkable maneuverability in 3D space. Ironically, object detection for drones has always been conducted in the 2D image space, which fundamentally limits their ability to understand 3D scenes. Furthermore, existing 3D object detection methods developed for autonomous driving cannot be directly applied to drones due to the lack of deformation modeling, which is essential for the distant aerial perspective with sensitive distortion and small objects. To fill the gap, this work proposes a dual-view detection system named DVDET to achieve aerial monocular object detection in both the 2D image space and the 3D physical space. To address the severe view deformation issue, we propose a novel trainable geo-deformable transformation module that can properly warp information from the drone's perspective to the BEV. Compared to the monocular methods for cars, our transformation includes a learnable deformable network for explicitly revising the severe deviation. To address the dataset challenge, we propose a new large-scale simulation dataset named AM3D-Sim, generated by the co-simulation of AirSIM and CARLA, and a new real-world aerial dataset named AM3D-Real, collected by DJI Matrice 300 RTK, in both datasets, high-quality annotations for 3D object detection are provided. Extensive experiments show that i) aerial monocular 3D object detection is feasible; ii) the model pre-trained on the simulation dataset benefits real-world performance, and iii) DVDET also benefits monocular 3D object detection for cars. To encourage more researchers to investigate this area, we will release the dataset and related code in https://github.com/PhyllisH/DVDET.
Autonomous bin picking poses significant challenges to vision-driven robotic systems given the complexity of the problem, ranging from various sensor modalities, to highly entangled object layouts, to diverse item properties and gripper types. Existing methods often address the problem from one perspective. Diverse items and complex bin scenes require diverse picking strategies together with advanced reasoning. As such, to build robust and effective machine-learning algorithms for solving this complex task requires significant amounts of comprehensive and high quality data. Collecting such data in real world would be too expensive and time prohibitive and therefore intractable from a scalability perspective. To tackle this big, diverse data problem, we take inspiration from the recent rise in the concept of metaverses, and introduce MetaGraspNet, a large-scale photo-realistic bin picking dataset constructed via physics-based metaverse synthesis. The proposed dataset contains 217k RGBD images across 82 different article types, with full annotations for object detection, amodal perception, keypoint detection, manipulation order and ambidextrous grasp labels for a parallel-jaw and vacuum gripper. We also provide a real dataset consisting of over 2.3k fully annotated high-quality RGBD images, divided into 5 levels of difficulties and an unseen object set to evaluate different object and layout properties. Finally, we conduct extensive experiments showing that our proposed vacuum seal model and synthetic dataset achieves state-of-the-art performance and generalizes to real world use-cases.
Recently CNN-based RGB-D salient object detection (SOD) has obtained significant improvement on detection accuracy. However, existing models often fail to perform well in terms of efficiency and accuracy simultaneously. This hinders their potential applications on mobile devices as well as many real-world problems. To bridge the accuracy gap between lightweight and large models for RGB-D SOD, in this paper, an efficient module that can greatly improve the accuracy but adds little computation is proposed. Inspired by the fact that depth quality is a key factor influencing the accuracy, we propose an efficient depth quality-inspired feature manipulation (DQFM) process, which can dynamically filter depth features according to depth quality. The proposed DQFM resorts to the alignment of low-level RGB and depth features, as well as holistic attention of the depth stream to explicitly control and enhance cross-modal fusion. We embed DQFM to obtain an efficient lightweight RGB-D SOD model called DFM-Net, where we in addition design a tailored depth backbone and a two-stage decoder as basic parts. Extensive experimental results on nine RGB-D datasets demonstrate that our DFM-Net outperforms recent efficient models, running at about 20 FPS on CPU with only 8.5Mb model size, and meanwhile being 2.9/2.4 times faster and 6.7/3.1 times smaller than the latest best models A2dele and MobileSal. It also maintains state-of-the-art accuracy when even compared to non-efficient models. Interestingly, further statistics and analyses verify the ability of DQFM in distinguishing depth maps of various qualities without any quality labels. Last but not least, we further apply DFM-Net to deal with video SOD (VSOD), achieving comparable performance against recent efficient models while being 3/2.3 times faster/smaller than the prior best in this field. Our code is available at https://github.com/zwbx/DFM-Net.