Saliency Detection with Moving Camera via Background Model Completion

Detecting saliency in videos is a fundamental step in many computer vision systems. Saliency is the significant target(s) in the video. The object of interest is further analyzed for high-level applications. The segregation of saliency and the background can be made if they exhibit different visual cues. Therefore, saliency detection is often formulated as background subtraction. However, saliency detection is challenging. For instance, dynamic background can result in false positive errors. In another scenario, camouflage will result in false negative errors. With moving cameras, the captured scenes are even more complicated to handle. We propose a new framework, called saliency detection via background model completion (SD-BMC), that comprises a background modeler and a deep learning background/foreground segmentation network. The background modeler generates an initial clean background image from a short image sequence. Based on the idea of video completion, a good background frame can be synthesized with the co-existence of changing background and moving objects. We adopt the background/foreground segmenter, which was pre-trained with a specific video dataset. It can also detect saliency in unseen videos. The background modeler can adjust the background image dynamically when the background/foreground segmenter output deteriorates during processing a long video. To the best of our knowledge, our framework is the first one to adopt video completion for background modeling and saliency detection in videos captured by moving cameras. The F-measure results, obtained from the pan-tilt-zoom (PTZ) videos, show that our proposed framework outperforms some deep learning-based background subtraction models by 11% or more. With more challenging videos, our framework also outperforms many high-ranking background subtraction methods by more than 3%.

Generic derivations on o-minimal structures

Let [Formula: see text] be a complete, model complete o-minimal theory extending the theory [Formula: see text] of real closed ordered fields in some appropriate language [Formula: see text]. We study derivations [Formula: see text] on models [Formula: see text]. We introduce the notion of a [Formula: see text]-derivation: a derivation which is compatible with the [Formula: see text]-definable [Formula: see text]-functions on [Formula: see text]. We show that the theory of [Formula: see text]-models with a [Formula: see text]-derivation has a model completion [Formula: see text]. The derivation in models [Formula: see text] behaves “generically”, it is wildly discontinuous and its kernel is a dense elementary [Formula: see text]-substructure of [Formula: see text]. If [Formula: see text], then [Formula: see text] is the theory of closed ordered differential fields (CODFs) as introduced by Michael Singer. We are able to recover many of the known facts about CODF in our setting. Among other things, we show that [Formula: see text] has [Formula: see text] as its open core, that [Formula: see text] is distal, and that [Formula: see text] eliminates imaginaries. We also show that the theory of [Formula: see text]-models with finitely many commuting [Formula: see text]-derivations has a model completion.

OPTIMASI PENDISTRIBUSIAN PRODUK MENGGUNAKAN MODEL FUZZY MULTIOBJEKTIF CYCLICAL INVENTORY ROUTING PROBLEM

In distribution activities, travel time is influenced by several factors, such as vehicle conditions, traffic conditions and road conditions. Uncertain traffic conditions cause the travel time can not be defined with certainty (uncertain). If several variables are in a uncertain condition then a deterministic approach is not appropriate. Fuzzy approach, can be used to overcome uncertainty conditions. In this study, the optimal route for chicken egg delivery using the fuzzy Multiobjective Cyclical Inventory Routing Problem (FMOCIRP) model with time and cost is expressed by triangular fuzzy numbers. Model completion using Winqsb software. An optimal shipping route is obtained with a total cost of Rp 114,000 and a total delivery time of 132 minutes. The first optimal delivery route is distributor, retailer 1, retailer 2, retailer 3 and back to the distributor. The second route is distributor, retailer 5, retailer 4 and back to the distributor.

METHODOLOGY AND METHODOLOGICAL APPROACHES TO STRUCTURING THE ORGANIZATIONAL AND ECONOMIC MECHANISM OF INNOVATIONORIENTED PUBLIC-PRIVATE PARTNERSHIP

The organizational and economic mechanism of innovationoriented public-private partnership should have a scientifically sound structure and a clear clearly structured conceptual model that can provide a superstructure for the regulation of its activities and development. It is provided by structuring and modeling of the organizational and economic mechanism is in accordance with the defined goals and innovation. That is, the structuring of the organizational and economic mechanism of innovation-oriented publicprivate partnership has a model completion and should have instrumental and functional support. The structuring of the organizational and economic mechanism of innovation-oriented public-private partnership in accordance with the defined goals, innovative technological solutions, created innovative value and other features reflects several of its variable models, which differ in format, but provide a holistic view of the mechanism. That is, the structuring of the organizational and economic mechanism of innovation-oriented public-private partnership has a model completion and should have instrumental and functional support.

Fast energy minimization of the CCDC drug-subset structures by molecule-in-cluster computations allows independent structure validation and model completion

Optimizing structures with computations on clusters of molecules permits generation of structure-specific restraints for refinement and structure validation.

Model theory of Steiner triple systems

A Steiner triple system (STS) is a set [Formula: see text] together with a collection [Formula: see text] of subsets of [Formula: see text] of size 3 such that any two elements of [Formula: see text] belong to exactly one element of [Formula: see text]. It is well known that the class of finite STS has a Fraïssé limit [Formula: see text]. Here, we show that the theory [Formula: see text] of [Formula: see text] is the model completion of the theory of STSs. We also prove that [Formula: see text] is not small and it has quantifier elimination, [Formula: see text], [Formula: see text], elimination of hyperimaginaries and weak elimination of imaginaries.

EXISTENTIALLY CLOSED BROUWERIAN SEMILATTICES

The variety of Brouwerian semilattices is amalgamable and locally finite; hence, by well-known results [19], it has a model completion (whose models are the existentially closed structures). In this article, we supply a finite and rather simple axiomatization of the model completion.

Impact of Chaining Method and Level of Completion on Accuracy of Function Structure-Based Market Price Prediction Models

The goal of this paper is to explore how different modeling approaches for constructing function structure models and different levels of model completion affect the ability to make inferences (reason) on the resulting information within the respective models. Specifically, the function structure models are used to predict market prices of products, predictions that are then compared based on their accuracy and precision. This work is based on previous studies on understanding how function modeling and the use of topological information from design graphs can be used to predict information with historical training. It was found that forward chaining was the least favorable chaining type irrespective of the level of completion, whereas the backward-chaining models performed relatively better across all completion levels. Given the poor performance of the nucleation models at the highest level of completion, future research must be directed toward understanding and employing the methods yielding the most accuracy. Moreover, the results from this simulation-based study can be used to develop modeling guidelines for designers or students, when constructing function models.