VisualVital: An observation model for multiple sections of scenes

Visual tracking is a challenging computer vision task due to the significant observation changes of the target. By contrast, the tracking task is relatively easy for humans. In this article, we propose a tracker inspired by the cognitive psychological memory mechanism, which decomposes the tracking task into sensory memory register, short-term memory tracker, and long-term memory tracker like humans. The sensory memory register captures information with three-dimensional perception; the short-term memory tracker builds the highly plastic observation model via memory rehearsal; the long-term memory tracker builds the highly stable observation model via memory encoding and retrieval. With the cooperative models, the tracker can easily handle various tracking scenarios. In addition, an appearance-shape learning method is proposed to update the two-dimensional appearance model and three-dimensional shape model appropriately. Extensive experimental results on a large-scale benchmark data set demonstrate that the proposed method outperforms the state-of-the-art two-dimensional and three-dimensional trackers in terms of efficiency, accuracy, and robustness.

Download Full-text

Hydro-climatic changes and their monitoring in the Arctic: Observation-model comparisons and prioritization options for monitoring development

Journal of Hydrology ◽

10.1016/j.jhydrol.2013.04.003 ◽

2013 ◽

Vol 492 ◽

pp. 273-280 ◽

Cited By ~ 15

Author(s):

Arvid Bring ◽

Georgia Destouni

Keyword(s):

Climatic Changes ◽

The Arctic ◽

Observation Model ◽

Model Comparisons

Download Full-text

Correcting observation model error in data assimilation

Chaos An Interdisciplinary Journal of Nonlinear Science ◽

10.1063/1.5087151 ◽

2019 ◽

Vol 29 (5) ◽

pp. 053102 ◽

Cited By ~ 1

Author(s):

Franz Hamilton ◽

Tyrus Berry ◽

Timothy Sauer

Keyword(s):

Data Assimilation ◽

Model Error ◽

Observation Model

Download Full-text

Conditional and Unconditional Deterministic Bounds on the MSE of the Non-Uniform Linear Co-centered Orthogonal Loop and Dipole Array

ENP Engineering Science Journal ◽

10.53907/enpesj.v1i1.11 ◽

2021 ◽

Vol 1 (1) ◽

pp. 13-20

Author(s):

Tao Bao ◽

Mohammed Nabil EL KORSO

Keyword(s):

Theoretical Analysis ◽

Closed Form ◽

Mean Square Error ◽

Source Localization ◽

Threshold Region ◽

Observation Model ◽

Mean Square ◽

Numerical Examples ◽

Cramer Rao Bound ◽

Deterministic Bounds

The co-centered orthogonal loop and dipole (COLD) array exhibits some interesting properties, which makes it ubiquitous in the context of polarized source localization. In the literature, one can find a plethora of estimation schemes adapted to the COLD array. Nevertheless, their ultimate performance in terms the so-called threshold region of mean square error (MSE), have not been fully investigated. In order to fill this lack, we focus, in this paper, on conditional and unconditional bounds that are tighter than the well known Cramér-Rao Bound (CRB). More precisely, we give some closed form expressions of the McAulay-Hofstetter, the Hammersley-Chapman-Robbins, the McAulaySeidman bounds and the recent Todros-Tabrikian bound, for both the conditional and unconditional observation model. Finally, numerical examples are provided to corroborate the theoretical analysis and to reveal a number of insightful properties.

Download Full-text

Sistem Pengamatan Kondisi Lalu Lintas Berbasis Data GPS pada Smartphone (Studi Kasus : Kota Denpasar)

Jurnal Teknologi Informasi dan Ilmu Komputer ◽

10.25126/jtiik.2020741909 ◽

2020 ◽

Vol 7 (4) ◽

pp. 667

Author(s):

Gede Herdian Setiawan ◽

I Ketut Dedy Suryawan

Keyword(s):

Real Time ◽

Traffic Congestion ◽

Data Retrieval ◽

Traffic Information ◽

Vehicle Speed ◽

Observation Model ◽

Economic Activities ◽

Traffic Condition ◽

Traffic Conditions ◽

Real Time Traffic

Pertumbuhan jumlah kendaraan yang semakin meningkat setiap tahunnya mengakibatkan volume kendaraan yang melintasi ruas jalan semakin padat yang kerap mengakibatkan kemacetan lalu lintas. Kemacetan lalu lintas dapat menjadi beban biaya yang signifikan terhadap kegiatan ekonomi masyarakat. Informasi lalu lintas yang dinamis seperti informasi kondisi lalu lintas secara langsung (real time) akan membantu mempengaruhi aktivitas masyarakat pengguna lalu lintas untuk melakukan perencanaan dan penjadwalan aktivitas yang lebih baik. Penelitian ini mengusulkan model pengamatan kondisi lalu lintas berbasis data GPS pada smartphone, untuk informasi kondisi lalu lintas secara langsung. GPS Receiver pada smartphone menghasilkan data lokasi secara instan dan bersifat mobile sehingga dapat digunakan untuk pengambilan data kecepatan kendaraan secara langsung. Kecepatan kendaraan diperoleh berdasarkan jarak perpindahan koordinat kendaraan dalam satuan detik selanjutnya di konversi menjadi satuan kecepatan (km/jam) kemudian data kecepatan kendaraan di proses menjadi informasi kondisi lalu lintas. Secara menyeluruh model pengamatan berfokus pada tiga tahapan, yaitu akuisisi data kecepatan kendaraan berbasis GPS pada smartphone, pengiriman data kecepatan dan visualisasi kondisi lalu lintas berbasis GIS. Pengujian dilakukan pada ruas jalan kota Denpasar telah mampu mendapatkan data kecepatan kendaraan dan mampu menunjukkan kondisi lalu lintas secara langsung dengan empat kategori keadaan lalu lintas yaitu garis berwarna hitam menunjukkan lalu lintas macet dengan kecepatan kendaraan kurang dari 17 km/jam, merah menunjukkan padat dengan kecepatan kendaraan 17 km/jam sampai 27 km/jam, kuning menunjukkan sedang dengan kecepatan kendaraan 26 km/jam sampai 40 km/jam dan hijau menunjukkan lancar dengan kecepatan kendaraan diatas 40 km/jam. AbstractThe growth in the number of vehicles that is increasing every year has resulted in the volume of vehicles crossing the road increasingly congested which often results in traffic congestion. Traffic congestion can be a significant cost burden on economic activities. Dynamic traffic information such as information on real time traffic conditions will help influence the activities of the traffic user community to better plan and schedule activities. This study proposes a traffic condition observation model based on GPS data on smartphones, for information on real time traffic conditions. The GPS Receiver on the smartphone produces location and coordinate data instantly and is mobile so that it can be used for direct vehicle speed data retrieval. Vehicle speed is obtained based on the displacement distance of the vehicle's coordinates in units of seconds and then converted into units of speed (km / h), the vehicle speed data is then processed into information on traffic conditions. Overall, the observation model focuses on three stages, namely GPS-based vehicle speed data acquisition on smartphones, speed data delivery and visualization of GIS-based traffic conditions. Tests carried out on the Denpasar city road segment have been able to obtain vehicle speed data and are able to show traffic conditions directly with four categories of traffic conditions, namely black lines indicating traffic jammed with vehicle speeds of less than 17 km / h, red indicates heavy with speed vehicles 17 to 27 km / h, yellow indicates medium speed with vehicles 26 km/h to 40 km / h and green shows fluent with vehicle speeds above 40 km / h.

Download Full-text

A sparse observation model to quantify species interactions in time and space

10.1101/815027 ◽

2019 ◽

Author(s):

Sadoune Ait Kaci Azzou ◽

Liam Singer ◽

Thierry Aebischer ◽

Madleina Caduff ◽

Beat Wolf ◽

...

Keyword(s):

Species Interactions ◽

Activity Patterns ◽

Niche Partitioning ◽

Camera Trap ◽

Camera Traps ◽

Observation Model ◽

Occupancy Models ◽

Time And Space ◽

Imperfect Detection ◽

Mobile Species

SummaryCamera traps and acoustic recording devices are essential tools to quantify the distribution, abundance and behavior of mobile species. Varying detection probabilities among device locations must be accounted for when analyzing such data, which is generally done using occupancy models. We introduce a Bayesian Time-dependent Observation Model for Camera Trap data (Tomcat), suited to estimate relative event densities in space and time. Tomcat allows to learn about the environmental requirements and daily activity patterns of species while accounting for imperfect detection. It further implements a sparse model that deals well will a large number of potentially highly correlated environmental variables. By integrating both spatial and temporal information, we extend the notation of overlap coefficient between species to time and space to study niche partitioning. We illustrate the power of Tomcat through an application to camera trap data of eight sympatrically occurring duiker Cephalophinae species in the savanna - rainforest ecotone in the Central African Republic and show that most species pairs show little overlap. Exceptions are those for which one species is very rare, likely as a result of direct competition.

Download Full-text

Validation of BeiDou Observations

Journal of Applied Geodesy ◽

10.1515/jag-2013-0027 ◽

2014 ◽

Vol 8 (2) ◽

Cited By ~ 1

Author(s):

Ahmed El-Mowafy ◽

Congwei Hu

Keyword(s):

Single Point ◽

Real Data ◽

Empirical Method ◽

Observation Model ◽

Triple Frequency ◽

Local Modeling ◽

Detection And Identification ◽

Using Data ◽

Navigation Information ◽

Graphical Diagnostics

AbstractThis study presents validation of BeiDou measurements in un-differenced standalone mode and experimental results of its application for real data. A reparameterized form of the unknowns in a geometry-free observation model was used. Observations from each satellite are independently screened using a local modeling approach. Main advantages include that there is no need for computation of inter-system biases and no satellite navigation information are needed. Validation of the triple-frequency BeiDou data was performed in static and kinematic modes, the former at two continuously operating reference stations in Australia using data that span two consecutive days and the later in a walking mode for three hours. The use of the validation method parameters for numerical and graphical diagnostics of the multi-frequency BeiDou observations are discussed. The precision of the system’s observations was estimated using an empirical method that utilizes the characteristics of the validation statistics. The capability of the proposed method is demonstrated in detection and identification of artificial errors inserted in the static BeiDou data and when implemented in a single point positioning processing of the kinematic test.

Download Full-text