Category: CSE Seminar Topics

Introduction to Artificial Intelligence and Global Risk:

Artificial Intelligence (AI) 

Since Charles Babbage has invited computer’s primitive form, human race have been technologically advancing. Many scientists and mathematicians have been trying to make computers think, behave and intelligent like humans. Mechanized intelligence is an advanced field of study and it includes both knowledge manipulation and representation.

In knowledge manipulation the main concept to solve algorithmic problems is that of search. The field of search should be explored only within a limit, this is called Heuristics. The ability to process new language is also part of artificial intelligence. Knowledge representation is made first in human mind and then only other steps will follow. This one deals with more of assumptions and predictions, the ability of thinking is a special feature of human brain and artificial intelligence should support languages based on various calculus such as lambda and predicate. These languages will help in various manipulations of data. 

 Global Risk 

The biggest risk that a software company faces is whether the project they have currently undertaken will be a failure or a success. The success of a project may rely on many factors such as proper planning, work allotment, completion on time, good testing and troubleshooting, excellent after sale service etc. The risk management consists of many stages such as problem, mitigation, prevention, anticipation, opportunity etc.

All these stages are inter-related with factors such as competition, customer satisfaction, predictability and numerous opportunities. It is pretty tedious job to implement these risk management in a company environment as there would be shortage of necessary tools for the analysis. The concept of seeing opportunity amongst chance of failure is called positive risk; this may be business, operational or systemic. 

The main risks associated with global software industry across environments such as legal, competitive, economic and social: 

• Entrant’s risk 

• Supplier’s risk 

• Substitute’s risk 

• Buyer’s Risk

 Download Artificial Intelligence and Global Risk PPT.

Introduction to A Nonlinear Feature Extractor for Texture Segmentation Seminar Topic:

Texture segmentation is the process by which images are partitioned into various different and distinct regions or areas. It is extremely useful in many areas such as surface inspection, image processing from satellites, picture archiving, image retrieval using content base etc.

Texture analysis techniques are classified into:

• Statistical
• Structural
• Transform based
• Model Based

In the common texture classification and segmentation process the input image is first transformed and then extracted in various stages. The extracted image is then conditioned and subjected to classification process. After classification we get the segmented image as the output. Various materials such as Gabor filters and wavelets are used the transformation process whereas smoothing filter and a transformation nonlinear functionis used for conditioning.

Finally before segmentation and during the classification stage, feature vectors are given as inputs to the classifier or a neural network which includes various feature images. In this article we are dealing with a network architecture used as an extractor which is nonlinear for the process of texture segmentation. These networks are 2-D and are feed forward. The advantages of these types of networks are it can have different variety of layers based during different texture segmentation. This neural model is represented by a mathematical expression.

The biggest advantage of this segmentation system is that it is independent of filters which are pre-defined. The texture images used for this system are the Brodatz textured images. The feature maps relates to the planes into which extracted layers are arranged, which is in turn distinguished by weight sets.

These weight sets are also known as mask sets will be distributed equally between all units of the map. These weights undergo training by specific set of algorithms which are finally used in the form of convolutional kernels. The pixel approach is always better than a region approach and there is less chance of errors.

 Download A Nonlinear Feature Extractor for Texture Segmentation Seminar Report.

Introduction to A Model-Based Approach to Evaluation of the Efficacy of FEC Coding in Combating Network Packet Losses:

During data transmission via network the loss of packets of data is one of the most common bugs we come across. FEC Coding or forward error correction coding ensures these errors are under control during the transmission of data over noisy and uncontrollable channels. Based onFEC, the receiver would be able to rectify the errors without needing to re send the data again.

The information is generally stored in a mass storage mechanism for easy retrieval. Main types of FEC coding are the block and convectional codes. Based on a particular framework a procedure is developed by evaluation session scenarios and the packet loss is evaluated based on it. In order to increase the performance of the system, the depths are interleaved, coding rates are channeled and lengths are blocked.

In the present system the loss of packets of data are assumed to be independent of the single multiplexer model process which subsequently will make the FEC performance notes estimation incorrect; moreover the results we get in numerical analysis with and without writing codes are totally different. Also the complexity is bit on the higher side during multi session. In the present system the correct efficiency of FEC cannot be judged at a correct proportion. In the proposed system a model based approach is proposed making it calculate the efficiency of FEC at more accuracy.

The single and multi-session conditions are taken and studied as separate ones reducing the overall complexity of the system. The losses occurred during the packet data transfer can be restored or recovered in a network on one condition that the rate and parameters of coding are chosen correctly. Due to this the efficiency of the performance of FEC can be noted and interpreted accurately. The complexity surrounding FEC and the subsequent calculation are considerably reduced.

 Download A Model-Based Approach to Evaluation of the Efficacy of FEC Coding in Combating Network Packet Losses.

Introduction to Artificial Passenger Seminar Topic:

The biggest problem for a driver while driving alone is that drowsiness can affect easily leading to fatal accidents and subsequent injuries. The proposed system is designed in such a way that the driver of a car would be in constant awake stage through a virtual passenger which chats with the driver.

The system tracks the face movement and various organ movements such as eyes, lips etc. By analyzing the face of the driver the application can give out a warning signal through the car stereo speakers and manage the driver to engage in a conversation. For that firstly we have make our voice profile and save it in the system. During driving when the application asks question, it will detect the variation in voice by comparing with the existing sound profile and warns us if we are feeling sleepy or we have a lack of concentration. 

The time taken for response is also noted by the system and if the response is swift and clear from the driver, a set of questionnaire will follow which the driver can engage with. This application can make funny and meaningful conversations regarding any topic such as music, news, weather, traffic etc. All these conversations are managed by a dialogue manager which inter operates with other systems inside the car such as stereo, GPS navigation, telephone directory etc. The software giant IBM developed a speech system called conversational interactivity for telematics especially for alone drivers, inside this system will be a natural language unit which manages the language of interaction.

A NLU is installed as embedded chipset inside the car and will have various grammar files to make it as authentic as possible. The response of the DU would be spontaneous. There would be an option to record and playback conversations done between the driver and the DU. If the system experiences driver fatigue it can alert or warn the driver to stop the car or lessen the speed.

 Download Artificial Passenger Seminar Report.

MANET is the abbreviation for mobile ad-hoc networks which in turn refers to a wireless network of mobile devices which can configure all by it and move independently. As each country has many powerful leaders, in MANET nodes the one which is having most resources left are nominated as leaders. This would help in balancing other nodes in the case of consumption of resources. All this nomination and election are using the Radial Basis function or in short RBF.

MANET’s lack a constant bottleneck; as a result the nodes in it have their own Intrusion detection system (IDS) to run which is a serious short coming of the system. In order to improve the above problem, leaders are selected from nodes for the entire cluster. The IDS is used to detect various threats coming towards the system, the network events are monitored and analyzed for any possible intrusions. The nodes are prevented from behaving selfishly and the nodes are allowed to include in the election process by revealing their sensitive information. 

No IDS detection is done in the present system, but in proposed system IDS is launched using RBF. This is done after leader selection using the cluster independent leader election or CLIE method. A particular set of mathematical equationsare conceived for this particular radical function. Each node and IDS has a unique set of identity and goals to achieve. The control of selfish and malicious nodes are extremely important and a mechanism should be there to prevent them being elected as the leader. The nodes must also be barred from releasing fake information in order to enhance their own reputation and benefits.

Some nodes may change their behavior once elected and deviate from their original work path and not offer IDS to nodes which are selected, such nodes must be controlled as well. To compute the valuation function cost of analysis function is important and also reputation of the system must be high. A specific algorithm is used to select the leader.

 Download A Dynamic Security Mechanism of Leader Election for Intrusion Detection in MANET Using RBF.

Introduction to Mobile Computing Seminar Topic:

We cannot imagine a world without computers or mobile phones especially the latter. Mobile phones are integrated in to our day to day lives as much as watches are. Nowadays we can use internet too to make long distant calls and this is called internet telephony.

This topic discusses various changes and advancement in the field of mobile technology, its limitations and future architectures and services like PAN and collaborative services. This topic also deals with vocal communication, SIM details, SMS, WAP, intelligent network etc.

Mobile computing involves:

• Mobile communication
• Mobile hardware
• Mobile software

When we throw the light on current mobile architecture the first thing that comes up is the voice communication which is the basic and primary objective of a mobile phone. Next up are the supplementary services such as forwarding the call, barring the call, sending the call to voice mail etc.

We can also enable or disable various things in the sim card such as moving or deleting contacts. There is hardly anyone who does not send a SMS; it is the basic text sending feature of a mobile phone. We can browse internet through WAP services and download applications to mobile made in Java or Symbian. In the case of advanced architecture, the splitting of logic and service content is extremely important, mobile services would be mostly multi-domain and there would also be Personal area networks for communicating between various devices. There are many types of network bandwidths such as low, variable or high bandwidth with corresponding latency.

There would be also heterogeneous networks which would be wireless networks. Another main feature of mobile computing is the seamless connectivity that it produces across many overlays. Low latency handoffs are implemented effectively leaving the system to be more authentic and secure. Nowadays there are many portable computing devices such as laptops, tablets, palmtops etc. Mobile computing also ensures the quality of connectivity among various bandwidths. 

 Download Mobile Computing Seminar Report.

Introduction to Microcontroller Based Trainer Kit:

Today almost all electronic gadgets or devices we use in day to day life are actually an assembled embedded system. They have been used everywhere ranging from a small clock to huge satellites. In these embedded system, there are few components such as microprocessors and microcontrollers which is programmed with codes.

Assembly programming forms the base of any embedded system although nowadays coders prefer languages such as Java or C++ as it is easier to execute. The pre-determined tasks are executed precisely by an embedded system; the tasks are set for each component in the embedded system. It may have various user interfaces, depend on CPU platform and use compilers; debuggers etc. while developing and are called with a startup process.

As I mentioned earlier microcontroller is an important component in the embedded system set up. There are many software and hardware components involved in the development of the system.

For setting up a microcontroller, one has to write the code first, then translate it and finally check it for any bugs. Embedded system helps us reduce energy consumption and confine the space required. The microcontroller we have used here is the 89c51 microcontroller.

Features of 89c51 controller:

• 8 bit microcontroller
• 128 MB RAM
• 4K ROM
• One Serial Port
• Stack Pointer 8 bit
• Transmitter/Receiver
• Data Pointer and program counter
• Timers and registers
• Circuits and interrupters
• Oscillator amplifier

The training kit we have used here consists of various components such as power adaptors, Sensors, LCD & LED’s, keyboards, converters, RTC etc. A trainer kit must be designed minimally, easy to use, flexible& versatile and must be cost effective.

It must also be able to handle multiple functions on one device. Various applications tested here are a digital clock, temperature monitoring system, wave generation etc. Many methods are used for data transmission including synchronous and asynchronous methods, full or half duplex method, simplex method.

 Download ECE Project Report on Microcontroller Based Trainer Kit.

Reference Links:

PPT on Microcontroller Based Projects

Introduction to Fluorescent Multilayer Disk Seminar Topic:

It was extremely difficult to store multimedia files till about 15-20 years ago. We relied heavily on tape or floppy disks which could hold only 1.44MB of data. Then was the compact disc or CD ROM revolution which was followed by DVD ROM. It was a huge step in data writing technology and was extremely helpful for people from all areas of life. While a CD could hold data up to 700MB, DVD can hold data up to 4.7GB.

The fluorescent multilayer disk or the FMD disk is basically a type of optical disk. Usually disks use reflective materials for data storage but FMD uses materials which are fluorescent in nature. Normally CD/DVD could be written only at two layers of it due to many reasons, but FMD disk could be written up to approximately 100 layers  with a data storage capacity of 1TB. FMD is a revolution in the disc technology and could be the one for the future.

The design of a FMD is bit different from a CD or DVD; it is completely transparent. There would be polycarbonate layers with structures called pits filled with photochromic fluorescent material are stacked over each other. The fluorescent material makes the coherent light incoherent by glowing off; these incoherent lights differ in wavelength.

Related Information:

Since the surfaces are transparent, the laser light gets filtered off resulting in a superior signal-noise ratio than a CD or DVD. This pushes the disc to have many numbers of layers.

The biggest advantage of this system is that the photo chrome which acts as the fluorescent material does not get worn out with time and as a result the longevity of the disc is assured. Since it is a different technology altogether FMD readers and writers would be also different from a CD-R or DVD-R. A new reading technique called parallel reading is used to read FMD-ROM discs.