As an instance-level recognition problem, re-identification (re-ID) requires models to capture diverse features. However, with continuous training, re-ID models pay more and more attention to the ...salient areas. As a result, the model may only focus on few small regions with salient representations and ignore other important information. This phenomenon leads to inferior performance, especially when models are evaluated on small inter-identity variation data. In this paper, we propose a novel network, Erasing-Salient Net (ES-Net), to learn comprehensive features by erasing the salient areas in an image. ES-Net proposes a novel method to locate the salient areas by the confidence of objects and erases them efficiently in a training batch. Meanwhile, to mitigate the over-erasing problem, this paper uses a trainable pooling layer P-pooling that generalizes global max and global average pooling. Experiments are conducted on two specific re-identification tasks (i.e., Person re-ID, Vehicle re-ID). Our ES-Net outperforms state-of-the-art methods on three Person re-ID benchmarks and two Vehicle re-ID benchmarks. Specifically, mAP / Rank-1 rate: 88.6% / 95.7% on Market1501, 78.8% / 89.2% on DuckMTMC-reID, 57.3% / 80.9% on MSMT17, 81.9% / 97.0% on Veri-776, respectively. Rank-1 / Rank-5 rate: 83.6% / 96.9% on VehicleID (Small), 79.9% / 93.5% on VehicleID (Medium), 76.9% / 90.7% on VehicleID (Large), respectively. Moreover, the visualized salient areas show human-interpretable visual explanations for the ranking results.
Recently, it is becoming a challenging work for person re-identification due to the problems of occlusion, blurring and posture. The key of effective person re-identification is to capture sufficient ...detailed features of a person's appearance in images. Different from previous methods, our method mainly focuses on fusing different visual clues only depending on the features of different levels and scales without additional assistance. The major contributions of our paper are the mixed pooling strategy with different kernels and the mixed loss function. Firstly, we adopt ResNet50 as our backbone. We have slightly modified the backbone, which does not use the down-sampling operation at the beginning of stage 4. Inspired by pyramid pooling structure, we pass the outputs of Res4 and Res5 through the average pooling layer and max pooling layer with different kernels and strides separately. Secondly, we combine the averaged triplet losses and the averaged softmax losses as the final loss of the whole network. Extensive experiments on three datasets (CUHK3, Market1501, DukeMTMC-reID) show that compared with many state-of-the-art methods in recent years, our model achieve higher accuracy.
This paper analytically examines the ride-pooling market equilibrium under a single or multiple ride-pooling service(s) considering different pooling sizes (number of riders sharing the same vehicle) ...and endogenous congestion effect of the ride-pooling fleet, and investigates the optimal operation decisions of the service operator(s) regarding the pricing, fleet size, and pooling size in different administrative and market regimes. We firstly examine the scenario with a single ride-pooling service platform given a specific vehicle type for either profit-maximization or social welfare-maximization (different administrative regimes), and then extend the model to the scenario with two differentiated ride-pooling services (given different vehicle types), where the platforms may either compete or cooperate (different market regimes). We have the following major findings. First, the pooling size should be optimized to accommodate the congestion effect and thus simultaneously increase operator’s profit and social welfare. The optimal pooling size balances the marginal pickup time, crowding cost and waiting time cost, and the optimal value is different in different market regimes. Second, the ride-pooling service operator will internalize the congestion effect of providing more vehicles to serve customers, which yields a larger ratio of waiting customers to vacant vehicles (larger demand–supply ratio). Under two differentiated ride-pooling services, in addition to the direct congestion effect of providing more vehicles, its effect on the other operator’s decision will also be internalized by the service operator in concern. Our results also suggest that the competition between platforms tends to reduce the waiting customers-vacant vehicles ratio (demand–supply ratio), while the cooperation between or integration of two operators might yield a large ratio. Third, when the endogenous congestion effect of fleet size is zero, the operator’s profit at social optimum will always be negative when the matching function exhibits increasing return to scale and the marginal operation cost is less than the average operation cost, which means that the social welfare seeking operator has to be subsidized. When the endogenous congestion effect of the fleet size is positive, the ride-pooling service operator may earn a positive profit at social optimum under certain conditions. This study enhances the understanding in relation to users’ reactions to differentiated ride-pooling services, how such services interact with congestion, and how such services can be optimized in different market regimes.
•Ride-pooling market equilibrium with flexible pooling size in the presence of congestion effect.•Optimization of pooling size, fleet size, and fare of ride-pooling service.•Differentiated ride-pooling services under competition/cooperation/integration market regimes.•Complex interactions among customer traveling behaviors, service operation decisions and road traffic congestion.
Deep neural networks are the most used machine learning systems in the literature, for they are able to train huge amounts of data with a large number of parameters in a very effective way. However, ...one of the problems that such networks face is overfitting. There are many ways to address the overfitting issue, one of which is regularization using the dropout function. The use of dropout has the benefit of using a combination of different networks in one architecture and preventing units from co-adapting in an excessive way. The dropout function is known to work well in fully-connected layers as well as in pooling layers. In this work, we propose a novel method called Mixed-Pooling-Dropout that adapts the dropout function with a mixed-pooling strategy. The dropout operation is represented by a binary mask with each element drawn independently from a Bernoulli distribution. Experimental results show that our proposed method outperforms conventional pooling methods as well as the max-pooling-dropout method with an interesting margin (0.926 vs 0.868) regardless of the retaining probability.
Deep convolutional neural networks (CNNs) have been successfully applied to a wide variety of problems in computer vision, including salient object detection. To accurately detect and segment salient ...objects, it is necessary to extract and combine high-level semantic features with low-level fine details simultaneously. This is challenging for CNNs because repeated subsampling operations such as pooling and convolution lead to a significant decrease in the feature resolution, which results in the loss of spatial details and finer structures. Therefore, we propose augmenting feedforward neural networks by using the multistage refinement mechanism. In the first stage, a master net is built to generate a coarse prediction map in which most detailed structures are missing. In the following stages, the refinement net with layerwise recurrent connections to the master net is equipped to progressively combine local context information across stages to refine the preceding saliency maps in a stagewise manner. Furthermore, the pyramid pooling module and channel attention module are applied to aggregate different-region-based global contexts. Extensive evaluations over six benchmark datasets show that the proposed method performs favorably against the state-of-the-art approaches.
•Dense connection and spatial pyramid pooling based YOLO (DC-SPP-YOLO) is proposed.•The DC-SPP-YOLO is developed for ameliorating the object detection accuracy of YOLOv2 by employing the dense ...connection structure of convolutional layers and introducing an improved spatial pyramid pooling.•The PASCAL VOC standard datasets and the UA-DETRAC vehicle datasets are used to verify the effectiveness of the proposed method.•Results show that the proposed method not only outperforms the YOLOv2 method but gets the state-of-the-art performance on object detection tasks.
Although the YOLOv2 method is extremely fast on object detection, its detection accuracy is restricted due to the low performance of its backbone network and the under-utilization of multi-scale region features. Therefore, a dense connection (DC) and spatial pyramid pooling (SPP) based YOLO (DC-SPP-YOLO) method is proposed in this paper for ameliorating the object detection accuracy of YOLOv2. Specifically, the backbone network of YOLOv2 adopts the dense connection of convolution layers, which strengthen the feature extraction and alleviate the vanishing-gradient problem. Moreover, an improved spatial pyramid pooling is introduced to pool and concatenate the multi-scale region features, so that the network learns the object features more comprehensively. The DC-SPP-YOLO model is established and trained based on a new loss function composed of MSE (mean square error) loss and cross-entropy loss. The experimental results indicate that the mAP (mean Average Precision) of DC-SPP-YOLO is higher than that of YOLOv2 on the PASCAL VOC datasets and the UA-DETRAC datasets. The effectiveness of DC-SPP-YOLO method proposed is demonstrated.
Omnichannel service delivery, which combines online and offline sales channels, is transforming many traditional service businesses with the advancement of IT technologies and the impact of the ...pandemic. This paper examines whether a service provider such as a restaurant should operate through dine-in only, delivery only, or omnichannel (i.e., using both channels simultaneously). For the omnichannel mode, we further compare a pooling system, where a single server handles both dine-in and delivery orders, and a dedicated system, where the restaurant uses separate servers to handle offline and online orders respectively. We focus on the restaurants that do not have their own online ordering systems and must rely on a third-party platform to receive and process online orders. The platform charges a commission fee to the restaurant for each online order. We model the interactions among the three parties: platforms, restaurants, and customers, as a Stackelberg game. To find the equilibrium, we first analyze the customer choice between online and offline channels, then study the restaurant’s decision on the channel choice and, if omnichannel is chosen, whether to adopt a pooling or dedicated system, and finally study the platform’s decision on the commission rate. We show that the pooling system yields a higher profit and throughput for the restaurant, while the dedicated system only generates a larger social welfare than the pooling system when the commission rate is relatively high. Interestingly, we reveal that under the dedicated system, the platform has an incentive to increase the commission rate to extract more profits from the delivery channel, while the pooling system can mitigate this platform’s opportunistic behavior.
Fusion of Probability Density Functions Koliander, Gunther; El-Laham, Yousef; Djuric, Petar M. ...
Proceedings of the IEEE,
04/2022, Volume:
110, Issue:
4
Journal Article
Peer reviewed
Open access
Fusing probabilistic information is a fundamental task in signal and data processing with relevance to many fields of technology and science. In this work, we investigate the fusion of multiple ...probability density functions (pdfs) of a continuous random variable or vector. Although the case of continuous random variables and the problem of pdf fusion frequently arise in multisensor signal processing, statistical inference, and machine learning, a universally accepted method for pdf fusion does not exist. The diversity of approaches, perspectives, and solutions related to pdf fusion motivates a unified presentation of the theory and methodology of the field. We discuss three different approaches to fusing pdfs. In the axiomatic approach, the fusion rule is defined indirectly by a set of properties (axioms). In the optimization approach, it is the result of minimizing an objective function that involves an information-theoretic divergence or a distance measure. In the supra-Bayesian approach, the fusion center interprets the pdfs to be fused as random observations. Our work is partly a survey, reviewing in a structured and coherent fashion many of the concepts and methods that have been developed in the literature. In addition, we present new results for each of the three approaches. Our original contributions include new fusion rules, axioms, and axiomatic and optimization-based characterizations; a new formulation of supra-Bayesian fusion in terms of finite-dimensional parametrizations; and a study of supra-Bayesian fusion of posterior pdfs for linear Gaussian models.
Lateral transshipments within an inventory system are stock movements between locations of the same echelon. These transshipments can be conducted periodically at predetermined points in time to ...proactively redistribute stock, or they can be used reactively as a method of meeting demand which cannot be satisfied from stock on hand. The elements of an inventory system considered, e.g. size, cost structures and service level definition, all influence the best method of transshipping. Models of many different systems have been considered. This paper provides a literature review which categorizes the research to date on lateral transshipments, so that these differences can be understood and gaps within the literature can be identified.
Two architectures that generalize convolutional neural networks (CNNs) for the processing of signals supported on graphs are introduced. We start with the selection graph neural network (GNN), which ...replaces linear time invariant filters with linear shift invariant graph filters to generate convolutional features and reinterprets pooling as a possibly nonlinear subsampling stage where nearby nodes pool their information in a set of preselected sample nodes. A key component of the architecture is to remember the position of sampled nodes to permit computation of convolutional features at deeper layers. The second architecture, dubbed aggregation GNN, diffuses the signal through the graph and stores the sequence of diffused components observed by a designated node. This procedure effectively aggregates all components into a stream of information having temporal structure to which the convolution and pooling stages of regular CNNs can be applied. A multinode version of aggregation GNNs is further introduced for operation in large-scale graphs. An important property of selection and aggregation GNNs is that they reduce to conventional CNNs when particularized to time signals reinterpreted as graph signals in a circulant graph. Comparative numerical analyses are performed in a source localization application over synthetic and real-world networks. Performance is also evaluated for an authorship attribution problem and text category classification. Multinode aggregation GNNs are consistently the best-performing GNN architecture.