Search

Bounding box estimation by overlap maximization has improved the state of the art of visual tracking significantly, yet the improvement in robustness and accuracy is restricted by the limited reference information, i.e., the initial target. In this paper, we present DCOM, a novel bounding box estimation method for visual tracking, based on distribution calibration and overlap maximization. We assume every dimension in the modulation vector follows a Gaussian distribution, so that the mean and the variance can borrow from those of similar targets in large-scale training datasets. As such, sufficient and reliable reference information can be obtained from the calibrated distribution, leading to a more robust and accurate target estimation. Additionally, an updating strategy for the modulation vector is proposed to adapt the variation of the target object. Our method can be built on top of off-the-shelf networks without finetuning and extra parameters. It yields state-of-the-art performance on three popular benchmarks, including GOT-10k, LaSOT, and NfS while running at around 40 FPS, confirming its effectiveness and efficiency.

Sequential patterns are important, because they can be exploited to improve the prediction accuracy of our classifiers. Sequential data, such as time series/video frames, and event data are becoming more and more ubiquitous in a wide spectrum of application scenarios especially in the background of large data and deep learning. However, large data sets used in training modern machine-learning models, such as deep neural networks, are often affected by label noise. Existing noisy learning approaches mainly focus on building an additional network to clean the noise or find a robust loss function. Few works tackle this problem by exploiting sample correlations. In this paper, we propose BundleNet, a framework of sequential structure (named bundle-module, see Fig. 1) for deep neural networks to handle the label noise. The bundle module naturally takes into account sample correlations by constructing bundles of samples class-by-class, and treats them as independent inputs. Moreover, we prove that the bundle-module performs a form of regularization, which is similar to dropout as regularization during training. The regularization effect endows the BundleNet with strong robustness to the label noise. Extensive experiments on public data sets prove that the proposed approach is effective and promising.

Efficient indexing and retrieving objects of interest from large‐scale surveillance videos are a significant and challenging topic. In this study, the authors present an effective multiple deep features learning approach for object retrieval in surveillance videos. Based on the discriminative convolutional neural network (CNN), they can learn multiple deep features to comprehensively describe the visual object. To be specific, they utilise the CNN model pre‐trained on ImageNet ILSVRC12 and fine‐tuned on our dataset to abstract structure information. In addition, they train another CNN model supervised by 11 colour names to deliver the colour information. To improve the retrieval performance, the deep features are encoded into short binary codes by locality‐sensitive hash and fused to fast retrieve the object of interest. Retrieval experiments are performed on a dataset of 100k objects extracted from multi‐camera surveillance videos. Comparison results with other common visual features show the effectiveness of the proposed approach.

In this paper, an improvement of deformable template matching algorithm for polygonal objects in grayscale images using two-dimensional deformable templates along orthogonal curves is presented. In the process of pre-computing extensions of the deformable template along orthogonal curves, the novel matching approach incorporates adapting knowledge-specific template discretization techniques appropriate for different polygonal objects and minimizing the improved internal and external energy terms containing inter-shape information of polygonal objects. In our application, this energy optimization problem of the deformable template is efficiently solved by a genetic algorithm (GA). Our algorithm has been successfully applied on synthetic images and real images. The experiment results show that the new approach provides more robust and accurate matching method.

We present an integrated framework on personalized sports video customization, which addresses three research issues: semantic video annotation, personalized video retrieval and summarization, and system adaptation. Sports video annotation serves as the foundation of the video customization system. To acquire detailed description of video content, external web text is adopted to align with the related sports video according to their semantic correspondence. Based on the derived semantic annotation, a user-participant multiconstraint 0/1 Knapsack model is designed to model the personalized video customization, which can unify both video retrieval and summarization with different fusion parameters. As a measure to make the system adaptive to the particular user, a social network based system adaptation algorithm is proposed to learn latent user preference implicitly. Both quantitative and qualitative experiments conducted on twelve broadcast basketball and football videos validate the effectiveness of the proposed method.

In this paper, we present an effective hierarchical shot classification scheme for broadcast soccer video. We first partition a video into replay and non-replay shots with replay logo detection. Then, non-replay shots are further classified into Long, Medium, Close-up or Out-field types with color and texture features based on a decision tree. We tested the method on real broadcast FIFA soccer videos, and the experimental results demonstrate its effectiveness..