|
|
An Efficient Algorithm for Small Gene Prediction in DNA Sequences
Hamidreza Saberkari, Rasht Branch Islamic Azad University, Iran
Mahsa Saffari Farsani, Yazd University, Iran
Sahar Aminkar, Tarbiat Modares University, Iran
Mousa Shamsi, Sahand University of Technology, Iran
Abstract:
The main purpose of this paper is to introduce a new method for gene prediction in DNA sequences based on the period-3 property in exons. First, the symbolic DNA sequences converted to digital signal by using maximum homogeny estimation modeling method. Then, to reduce the effect of background noise in the period-3 spectrum, we have used the discrete wavelet transform (DWT) at four levels and apply it on the input numerical strand. Finally, to extract the period-3 components in smoothed sequence, we have used the minimum variance spectrum estimating technique. Using the proposed algorithm leads to increase the speed of process and therefore to reduce the computational complexity. The ability of detect small size exons in DNA sequences is another advantage of our algorithm. Performance of the proposed algorithm in exon prediction is compared with several existing methods at the nucleotide level using: (i) specificity vs. sensitivity; (ii) receiver operating curves (ROC) curve; (iii) area under ROC curve. Simulation results show that our algorithm increase the accuracy of exon detection relative to the most common digital signal processing (DSP) tested methods for gene prediction.
Full Paper (in PDF)
|
PVBMT: A Principal Verb based Approach for English to Bangla Machine Translation
Masud Rabbani, Daffodil International University, Bangladesh
Kazi Md. Rokibul Alam, Muzahidul Islam, Khulna University of Engineering and Technology, Bangladesh
Yasuhiko Morimoto, Hiroshima University, Japan
Abstract:
This paper proposes principal verb based machine translation (PVBMT),
a new approach of machine translation (MT) from English to Bangla
(EtoB) that runs in both web-based and mobile applications. The key
mechanism is to detect the principal verb from any form of English
sentence and then to transform it into the simplest form of English
sentence i.e, subject plus verb plus object; identical to rule based MT
(RBMT). Also while a 'prepositional phrase (PP)' or an 'idiom and
phrase (I&P)' exists in a sentence, PVBMT uses its own corpus to
tag and bind it properly; identical to statistical MT (SMT). While
only RBMT is employed, often it generates feeble output because
it requires the matching of various forms of English sentences with
established grammatical rules stored in the knowledge-base. Therefore
PVBMT employs hybrid machine translation (HMT) paradigm, a hybrid
of RBMT and SMT. Finally the performance of PVBMT has been
compared with a number of existing on-line EtoB translators employing
a syntactic and a semantic analyzer. The experimental result shows that
PVBMT can translate any form of English sentence i.e., Interrogative,
Imperative, Exclamatory, Active, Passive, Complex or Compound along
with an 'I&P' or a 'PP' with better accuracy than others.
Full Paper (in PDF)
|
View Invariant Vehicle Type Recognition and Counting System using Multiple Features
Martins E. Irhebhude, Nigerian Defence Academy, Kaduna, Nigeria
Amin Nawahda, Sohar University, Oman
Eran A. Edirisinghe, Loughborough University, UK
Abstract:
This paper presents an automatic vehicle type recognition, classification
and counting system that is view invariant and can be used in real-time
road transportation and environmental pollution management systems.
A Support Vector Machine (SVM) classifier is used to classify vehicles
into four categories namely; cars, jeeps, buses and trucks, respectively,
based on measurable image features. Image analysis is performed on a set
of features that consists of Region, Histogram Oriented Gradient (HOG)
and Local Binary Pattern (LBP) histogram features. A feature combination
approach is proposed for recognition; Region, LBP and HOG
(RLH). Correlation based Feature Selection (CFS) is used to select the
most discriminative features from the above feature set thereby improving
recognition accuracy and reducing the time required for classification.
Various success rates are reported from the experiments conducted
on two separate datasets, with average accuracy reaching 95% on the
combined datasets. Proposed feature combination techniques compared
with Region, HOG, LBP, with results showing higher recognition accuracy
achievable when using the proposed feature combinations. Separate
frontal/rear and angular view datasets have been used in the experiments
to demonstrate the adaptability of the proposed algorithm to variations
of the view angle. Initially three practical application scenarios that will
benefit from the proposed technology are presented namely; access control,
toll collection and the estimation of air pollution levels caused by
vehicular traffic, justifying the practical importance and relevance of the
research presented.
Full Paper (in PDF)
|
Constrained Procedural Modeling of Real Buildings from Single Facade Layout
Divya Udayan J, VIT University, India
HyungSeok Kim, Konkuk University, South Korea
Abstract:
We present a novel modeling framework to generate real-world 3D buildings from a single facade layout that adapt to the real footprint data
automatically. Facade components are extracted from the facade layout and organized as a repetitive shape tree. Building facade layout
encodes expected architectural constraints and is able to derive complex
instances using shape grammars. We extend the previous approaches
of procedural building models to a constraint-based framework for the
recovery of the hidden parts of the building. We then provide an interactive editing process for updating of the structural topology given a
di erent view of the building. We demonstrate our framework on several
real-world buildings with varying amount of complexity in appearance
and footprint shape and we show that our approach can generate similar buildings to the original that are used to populate a virtual city
depicting implicit culture and style of di erent ages, while opening new
perspectives for image understanding and 3D modeling.
Full Paper (in PDF)
|
|
