Pedestrian Recognition with a Learned Metric

Date

2010-11-12

Author

DIKMEN, Mert
Akbaş, Emre
HUANG, Thomas S.
Ahuja, Narendra

Metadata

Show full item record

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

Item Usage Stats

123
views

0
downloads

This paper presents a new method for viewpoint invariant pedestrian recognition problem. We use a metric learning framework to obtain a robust metric for large margin nearest neighbor classification with rejection (i.e., classifier will return no matches if all neighbors are beyond a certain distance). The rejection condition necessitates the use of a uniform threshold for a. maximum allowed distance for deeming a, pair of images a match. In order to handle the rejection case, we propose a novel cost similar to the Large Margin Nearest Neighbor (LMNN) method and call our approach Large Margin Nearest Neighbor with Rejection (LMNN-R). Our method is able to achieve significant improvement over previously reported results on the standard Viewpoint Invariant Pedestrian Recognition (VIPeR [1]) dataset.

URI

https://hdl.handle.net/11511/55502

Collections

Department of Computer Engineering, Conference / Seminar

Suggestions

OpenMETU
Core

Frequency-domain subspace identification of linear time-periodic (LTP) systems Uyanık, İsmail; Saranlı, Uluç; Ankaralı, Mustafa Mert; Cowan, Noah J.; Morgül, Ömer (Institute of Electrical and Electronics Engineers (IEEE), 2019-06) This paper proposes a new methodology for subspace-based state-space identification for linear time-periodic (LTP) systems. Since LTP systems can be lifted to equivalent linear time-invariant (LTI) systems, we first lift input-output data from an unknown LTP system as if they were collected from an equivalent LTI system. Then, we use frequency-domain subspace identification methods to find the LTI system estimate. Subsequently. we propose a novel method to obtain a time-periodic realization for the estimate...
Traffic assignment in transforming networks case study : Ankara Zorlu, Fikret; Türel, Ali; Department of City and Regional Planning (2006) This study investigates the relevance of dynamic traffic assignment models under uncertainty. In the last years researchers have dealt with advanced traffic control systems since road provision is not regarded as a proper solution to relieve congestion. Dynamic assignment which is an essential component of investment planning is regarded as a new research area in the field of urban transportation. In this study the performance of dynamic traffic assignment method, which incorporates time dependent flow, is ...
State Space Representation of Smart Structures Under Unsteady Aerodynamic Loading Fatih Mutlu, Karadal; Güçlü, Seber; Şahin, Melin; Volkan, Nalbantoğlu; Yaman, Yavuz (null; 2007-09-10) This study presents a technique for the state space representation of the aeroelastic models of smart structures. It was based on a rational approximation of the unsteady aerodynamic loads in the Laplace domain, which yielded state-space matrix equations of motion with constant coefficients. In this study, an aluminum plate-like structure with twenty-four surface bonded piezoelectric patches was considered. The unsteady aerodynamic loads acting on the structure were calculated for a range of reduced freque...
DIRECTIONALLY ADAPTIVE SUPER-RESOLUTION Turgay, Emre; Akar, Gözde (2009-11-10) In this paper a novel direction adaptive super-resolution (SR) image reconstruction method is proposed. The proposed maximum a-posteriori (MAP) based estimator uses gradient direction for optimal noise reduction while preserving the edges. Compared to the other edge-preserving methods, the proposed algorithm uses gradient direction in addition to the gradient amplitude for optimum regularization. The method comprises a gradient amplitude and direction estimation stage where a gradient direction map is obtai...
Vision based obstacle detection and avoidance using low level image features Senlet, Turgay; Halıcı, Uğur; Department of Electrical and Electronics Engineering (2006) This study proposes a new method for obstacle detection and avoidance using low-level MPEG-7 visual descriptors. The method includes training a neural network with a subset of MPEG-7 visual descriptors extracted from outdoor scenes. The trained neural network is then used to estimate the obstacle presence in real outdoor videos and to perform obstacle avoidance. In our proposed method, obstacle avoidance solely depends on the estimated obstacle presence data. In this study, backpropagation algorithm on mult...

Citation Formats

M. DIKMEN, E. Akbaş, T. S. HUANG, and N. Ahuja, “Pedestrian Recognition with a Learned Metric,” 2010, vol. 6495, Accessed: 00, 2020. [Online]. Available: https://hdl.handle.net/11511/55502.