Resources

The Department of Computer Science and Engineering, Bangalore has the following active research labs which covers broader computer technology domains where the students, researchers, and faculty work to solve real-world problems.

• Eye Tracking & Computer Vision Research Lab
• Human Language Technology (HLT)
• Amrita Robotics Research Lab (ARRL)
• CAcHe: Computer Architecture and High-Performance Lab
• Artificial Intelligence Lab

The research facilities available in the above mentioned labs is listed below:

1	Desktops	Dell Optix	Human Language Technology Lab
2	Workstation	Dell Precision Tower 5810 XCTO Base
3	Workstation	Dell Precision Tower 5810 XCTO Base
4	Laptop	Mobile Precision 5530 CTO BASE
5	Multifunction printer	HP LAZERJET PRO MFP M128fn Printer
6	Digital Voice Recorder	SONY LCD UX560 Digital Voice Recorder
7	Storage Device	WD’Z TB Storage Device
8	Audio Mixer	ALLEN & HEATH ZED 10CHANNEL MIXER
9	Microphone	Seinheizer E835S
10	Headset with microphone	Plantronix USB478 stereo headset
11	Collar microphone	Lavalier Microphone MIC-J044
12	wireless headset	Wireless Headphone with Mic/ Balt Glint-BT06 ot
13	Voice Recorder	(Sony – ICD-UX560F)
14	AMD Graphic card -1 in Quantity	Tahiti Pro (AMD/ATI)	Computer Architecture and High Performance Computing Lab
15	NVIDIA Graphic card -8 in Quantity	NVIDIA Geforce GTX Titan Black (GK 110B)
16	NVIDIA Graphic card -4 in Quantity	NVIDIA GeForce GTX 750 (GM107)
17	NVIDIA Graphic card -9 in Quantity	NVIDIA GeForce GTX 760 (GK 104)
18	Processor- 8 in Quantity	Intel Core I7-4770 Processor (8GB DDR3 RAM)
19	Large Screen Monitors – 8 in Quantity	32″ Samsung Large Format Monitor Model No. MD32C
20	Processor and Monitor – 6 in Quantity	DELL OPTIPLEX 9020 and E1914Hc
21	ARM LPC2148 (25 Nos)	Microcontroller Board	Embedded Systems Lab
22	CPLD Board (25 Nos)
23	Spartan 3 (25 Nos)	FPGA Board
24	Spartan 6 FPGA Board (4 Nos)	FPGA Board
25	Viirtex 5 FPGA Board (1 No)	FPGA Board
26	Firebird V (10 Nos)	Mobile Robot	Robotics Lab
27	Bioloid Premium Kit (1 No)	Humanoid Robot
28	Multi-PIE, FIA and 3DFE Facial Expression Database (1 No)	Database
29	Wanhao Duplicator I3 Mini FDM 3d Printer Thermal Receipt Printer (3 nos)	3D Printer	AI Lab
30	3DPrintronics PRUSA i3 MK3 3D Printer(Assembled)	3D Printer
31	Electro Blocks Kit	Robot kit
32	Ranger Kit	Robot kit
33	Codey Kit	Robot kit
34	Quadrino Kit	Robot kit
35	Drone Starter Kit (7 nos)	Drone Kit
36	Advance kit for Arduino
37	Raspberry pi Kit
38	Laptop	Lenovo laptop	Eye Tracking & Computer Vision Research Lab
39	Eye Tracker	SMI RED n Professional
40	Eye Tracker	Eye Tribe

Embedded Systems Lab

The Embedded Systems lab supports teaching and research in embedded systems for PG and UG students. It is a well equipped lab with state of the art equipments. Equipments in the lab are ARM7 kits 25 in number, ARM9 kit, ARM CORTEX-M3, with peripherals like A/D converter, D/A converter, Elevator control, traffic light control, real time clock, stepper motor interface, DC motor interface, keypads, and LCD displays and lab is supported with 30 computer systems.

The lab is also equipped with FPGA kits for research in reconfigurable computing and for teaching purpose as well. It has CPLD kits 25 in number, Spartan3 kits 25 in number, Spartan6 kits 4 in number, and a Virtex5 kit. The lab also supports image processing research with a DEV 8000 evaluation kit and Embest university package.

Robotics is another research area the lab supports. The lab has a wheeled robot and an aqua bot which is an in-house built along with many mobile robots used for teaching purpose. The lab is supporting building of a Humanoid robot as part of its research activity.

The department of CSE has set up a few labs in order to carry out research in the identfied thrust areas. These labs have facilites to conduct research at all levels. Students from Undergraduate and Postgraduate programmes are actively involved in this lab to do research. Also research scholars and faculty members guide these students to do research in these labs. We aim at setting up a reseach hub that would eventually become a center of excellence.

Eye Tracking and Computer Vision Research Lab

About the Lab and Purpose

“We are what we see “the interdisciplinary Artificial Intelligence Research lab with specific focus in the area of Computer Vision, Medical Image Processing, Video and Eye Tracking Data Analytics. The lab supports specialized PG Electives of Computer Vision, Image Analysis etc. The lab projects focus in the area of health care and social use cases like Reading disability, Glaucoma detection, Parkinson Detection, monitoring mental health of a person, Assistance for ALS Patients, Deep Learning Research, Medical Image Fusion, Low cost eye gaze tracker, Damage Detection, Image to Text OCR etc., The Research Lab takes up consultancy works from industry.

Equipments

• SMI RED n Professional eye tracker
• Eye Tribe eye trackers

Software Loaded

The software required to analyze the eye tracking data

• Experimental suite
• Begaze

Research Data Collection

Standard Procedure

Steps of Data Recording using Eyetracker:

Data Recording is the pivotal point in all eye-tracker based research. Following are the steps of experimental set-up:

1. Test that any external equipment can interface with the eye tracker correctly.
2. Select the stimulus preparation and test them for timing.
3. Conduct pilot study to ensure the quality in technical set-up, instructions and data analysis, before recruiting real participants.
4. Treat participants so that they feel reasonably good about the task.
5. Have all consent forms, rewards, and other documentation prepared and copied before participants arrive.
6. Adjust eye camera view for a given participant for good data quality.
7. Look at the participant’s eyes to quickly identify any mascara, drooping eye lids, contact lenses, squint, or difficult glasses.
8. Conduct calibration of each participant.
9. Instruct your participants about the task, which should be precise and same to all participants.
10. After the data recording is finished, backup the data files.

Recording high-quality eye tracking data is a skill that must be learnt through training. Poor data quality can never be remedied by later data processing and statistical analysis, just like a good quality of your data can never compensate for an inadequate experimental design.

Internships

An internship titled ‘Bone growth analysis in 3D micro CT images” was offered to the B.Tech students from 3rd December 2018 to 14th December 2018 in collaboration with Amrita Institute of Medical Science, Kochi.

Research works completed in Eye Tracking and Computer Vision Research Lab

• Assistive communication Application using Eye Tracking for ALS Patients
• It is an assistive application for ALS patients in which a patient can convey their basic needs through Human Computer Interaction using a low cost Eye Tracking device. It has an uncomplicated and painless way to convey the basic needs of a patient. The application is developed in such a way that the care takers or doctors have the flexibility to modify the GUI according to each patient’s explicit needs.
• Eye Tracking in anomaly detection in medical images

Optic Disc (OD) detection is a key pre-processing component in many algorithms designed for automatic extraction of retinal anatomical structures and lesions thus, reliable and efficient OD localization is significant tasks in ophthalmic image processing. In this research a novel architecture for optic disc detection system using eye gaze tracking has been proposed. A patent has been published on system and detection of optic disc in fundus retinal image which can be extended for various other medical images.

• Patent:

SYSTEM AND METHOD FOR DETECTION OF FEATURES IN AN IMAGE USING KNOWLEDGE OF EXPERT EYE GAZE PATTERN, Application No.201641037789 A Publication Date 2018

• Detection of Stress using Wearable Sensors in IoT Platform
• The proposed work concentrates on developing an IoT system which can efficiently detect the stress level of a person and provide a feedback which can assist the person to cope with the stressors. The system consists of a smartband module and a chest strap module which can be worn around wrist and chest respectively.
• Eye Gaze as an Indicator for Stress Level Analysis in Student

It is a user friendly approach for analyzing the students’ stress level using an Eye Tracker. Student’s eye gaze parameters are recorded using an Eye Tracker. Various stimuli are used to analyze the increase in stress level with respect to the increase in cognitive load.

• Low Cost Eye Gaze Tracker Using Web Camera-
• Proposed a system which uses web camera and the free open source Computer Vision Library Open CV. The system captures real time video of the person to detect the eyes in the initial frames and extract the features of eyes. Once the features are extracted, the eyes are tracked in the subsequent frames and the gaze direction is estimated using the computational intelligence techniques.
• Navigation Through Eye-Tracking for Human–Computer Interface-
• Vision Buy is a productive tool in enhancing the user experience of E-commerce websites. The consumer has the ability to purchase anything at anytime from anywhere through their visual attention and eye movements. The process of analysis typically involves examining the characteristics and patterns of visual attention during the online shopping process. The selection of the product based on the consumer’s eye movements is done by adopting the principle of attention distribution, which refers to the percentage of time visually allocated to each category of product available on the web page. This data will be analyzed based on gaze points and based on these selections it enables users to navigate through the webpages. Thus, it is an ensuing product for any e-commerce web application.
• Third Eye: Assistance for Reading Disabiltiy-
• Third Eye uses eye gaze information from the eye tracker and analyzes it to extract features which can identify the reading skills of the subject and further broadly categorizes them based on their reading skills. An automated report is generated which depicts the areas of difficulty faced by the reader through various visualization techniques.
• Tracking Eye Movements To Predict The Valence of A Scene
• Studying human bio signals such as eye movements and tracking them can help in identifying and classifying the emotional essence of a scene. The existing methods employed to evaluate the reaction of the eyes based on exposure to a scene or image often use a classifier to extract features from eye movements. These extracted features are then evaluated to determine the valence of a scene.
• An Online Platform for Diagnosis of Children with Reading Disability
• A user friendly approach for analyzing the children’s dyslexic disorder level using an Eye Tracker. Proposed module uses set of input stimulus to analyze all the parameters that could identify the stage of disability in a child.
• Deep Learning Approach for Suspicious Activity Detection from Surveillance Video
• Deep learning approach is used to detect suspicious or normal activity in an academic environment, and which sends an alert message to the corresponding authority, in case of predicting a suspicious activity.
• Intelligent Gaze Tracking Approach for Trail Making Test
• Trail making test is a cognitive impairment test used for understanding the visual attention during the visual search task. The classical paper pencil method measures the completion time of the participant and there was no mechanism for comparison across the participant with similar feature. The psychologist has to observe the reactions of the participants during the trial process and there is no mechanism to capture it. This study made an attempt to resolve the above problem and tried to infer additional parameters which can support psychologist to understand the participant performance in trail making test.

Ongoing research work

• Mental Health Monitoring System using Eye Gaze Tracking

Develop an intelligent system which can automatically identify patterns and bio markers from the eye gaze of a person and helps to monitor their mental health. Machine Learning and Deep Learning technologies help the system to take Intelligent decision and can be widely used in health care domain to monitor the mental health of a person.

• Eye Gaze Based Recommendation system

Develop an intelligent system that can provide deeper insights into the performance of learners for an online assessment. The system will record the eye gaze patterns of the learner and evaluate the cognitive parameters that affect their performance. The system also predicts the skills required to achieve the learning goals.

• Eye Movement Analysis of Glaucoma Patients using Eye Tracker Device

Develop cost-effective HCI (Human Computer Interaction) platform which estimates visual field loss of participants during visual exploration task. The proposed system considers the state of art of eye gaze measures that checks restrictedness in the visual field.

• Analysis of eye tracking data to indicate visual cognition deficiency in PD patients.

PD patients shows visual impairment as an early sign. Predicting the level of visual cognition impairment through eye tracking data of PD patients helps the physicians to understand the cognitive impairments and treat the patient accordingly. A machine learning model is used to make predictions with an eye tracking framework.

Colloborations

PhD Courses offered

• CS802 Eye Tracking Methods and Measures
• 16CS711 Data Visualization

Dataset Created

• Eye Tracking Dataset-1
  1. Study on Stress Analysis–Participants: students from Amrita School of Engineering, Bengaluru
     No. Of Participants- 20
     Duration: 2 months
  2. Study on identifying the cognitive impairment. Dataset based on Trail Making Test, Participants: Employees from HCG, Number of participants: 30
  3. Study on emotional arousal – Dataset based on Pleasant and unpleasant images, Participants: Employees from HCG, Number of participants: 20
• Eye Tracking Dataset – 2
  1. Visual search task: 36 participants (31 PD patients and 5 controls) Eye tracking data
  2. Trail making Test -11 participants (9 PD patients and 2 controls) Eye tracking data
• Eye Tracking Dataset -3
  1. Debugging Taks-33 participants. (Data of software developers, testers, bids and proposal engineers, project managers etc). Study conducted at ABB
  2. Debugging task- 50 participants (4 experts, 28 learners and 10 novices)
  3. MCQ Test- 15 Participants
• Eye Tracking Dataset -4
  1. Study on Visual Exploration task – 60 participants (34 normal participants, 26 glaucoma patients). Duration: 3 months
• Eye Tracking Dataset –5
  1. Study on Radiologists gazing on Fundus Images – 15 Participants

Models

• Fuzzy based intelligent gaze tracking system to classify the participants based on their cognitive impairment.
• Fuzzy based intelligent system to predict the skills of the learners.
• Gaze Fusion-Deep Neural Network Model for Glaucoma Detection.
• Deep Learning Model for Text Detection in Images
• Eye Gazed Optic Disc Detection – MatLab Model
• FSTS – Fuzzy Based System for Target Search – A Matlab Model

CAcHe- Computer Architecture and High Performance Lab

About the Lab and Purpose

The CAcHe Lab was established in the year 2014. It is situated in fifth floor, E Block, Bangalore Campus. The lab is equipped with Intel Core I7 Processors along with the graphics cards such as Nvidia Geforce GTX (Titan Black/750/760) and AMD Tahiti Pro along with 32″ Samsung Large Screen Monitors. The research scholars are involved in the life science applications of computer aided drug discovery using HPC. Some of the research work focuses on designing of cache coherent techniques. The other direction of work is taking place in the area of High Level Synthesis.

The core objective of the lab is to support research in the following areas.

• Memory Design techniques in Multi-core architecture
• CPU/FPGA Hybrid Computing
• Exploration of massively parallel GPGPU architecture
• Exploring synergy among diverse devices in Heterogeneous computing systems
• Life science applications using HPC dealing with Large scale data mining of biological data. 
• Deep Learning

Equipments

HPC Lab contains 14 machines with Processor Intel Core I7 with a speed of 3.4 GHz along with 8 GB RAM. All theses machines are enabled with NVidia cards such as Geforce GTX 750 (512 CUDA Cores), Titan Black (2880 CUDA cores), GTX 760 (1152 CUDA cores) along with the machine having AMD card Tahiti Pro.

Software Loaded: Simulators and Frameworks

• Gem 5 Simulators High Level
• Synthesis Tools (Legup, Vivado)
• GPGPU-Sim simulator
• Visual Molecular Dynamics
• Schrödinger Maestr
• Gromacs-5.1.2
• Hadoop
• Hive
• Sqoop
• MySQL
• Matlab
• TensorFlow
• CUDA

Lab courses conducted

PG students develop their Deep Learning based projects in the lab. As such no lab course has been conducted there.

Ongoing projects for research labs

The PhD scholars utilize the lab for their research work.

Ongoing projects for research labs

• Designing and development of Cache memory techniques for Multi-core processor heterogeneous systems that includes cache coherence, cache partitioning schemes and low power cache designs. The architectural simulators such as Gem 5, Simple scalar are being used for experimentations.
• Enabling hybrid computing for CPU/FPGA. HLS tools such as Vivado is being used for the experimentation purpose.
• Exploring the massively parallel architectures using GPGPU architecture. To find a match between application level diversity and architectural features currently two approaches are being used in this regard: Using overlay on FPGA and using open source simulators. Tools such as Vivado and Verilator and GPGPU-Sim simulators are being used for the experimentation.
• Creating synergy among diverse devices in Heterogeneous computing system. CUDA framework is being used to execute benchmarks in nVidia GPU cards. HLS tools such as Vivado is being used for experimentation to run OpenCL benchmarks.
• Life science applications using HPC: Protein protein interactions are the building blocks of molecular processes in a cell. These interactions are being analyzed for identifying therapeutic targets. The database of available 3D-structures of the protein complexes are analyzed using Visual Molecular Dynamics (VMD-free version) and Tcl Scripting. Schrödinger Maestro (free version) is used for preparation of complexes. Gromacs-5.1.2 is used for Molecular Dynamic (MD) simulations of large biomolecules to be analyzed.

Artificial Intelligence Lab

Artificial intelligence lab was established in the year 2019. The lab supports teaching and research in AI for PG and UG students. It is a well equipped lab with state of the art equipments. The lab supports fast and easy prototyping with 3D printers.

Major Equipments List

Laboratory Courses Conducted

1. Introduction to Drones
2. Object Oriented Programming
3. Elements of Computing
4. Introduction to Digital Manufacturing
5. Introduction to Robotics
6. Intelligence in Biological Systems
7. Computational Engineering Mechanics
8. Robot Operating Systems and Robotics Simulator
9. Data Structures and Algorithm
10. Python for Machine Learning
11. Introduction to Communication Systems
12. Operating Systems

Cisco Networking Lab

The main purpose of the lab is to conduct Cisco Certification Courses

Equipments:

• Router -2811 (4 Numbers)
• Router -2811 (4 Numbers)
• Router -1800 (3 Numbers)
• CATALIST 3560 24 10/100 (2 Numbers)
• CATALIST 2560 24 10/100 (2 Numbers)
• CATALIST 2550 24 10/100 (2 Numbers)
• 6 AP WLAN Controller
• 16Port Asynchronous Module
• ASA5510 (2 Numbers)
• Photographs:
• Software Loaded:
• 2811Security Bundle
• Cisco Packet Tracer
• Lab courses conducted
• Routing and Switching labs

Embedded Systems Lab

Embedded Systems lab was established in the year 2012. The lab supports teaching and research in embedded systems for PG and UG students. It is a well equipped lab with state of the art equipments for embedded programming, FPGA programming and IoT application development.

Major Equipments List

Software Details

• Xilinx System Edition – 13
• EDK, System generator, Logic Edition
• Embest IDE for ARM
• Keil uVision 4
• RTOS – WinCE net 6.0 BSP, Linux 2.6.28 BSP, uC/OS-II and ucLinux

Laboratory Courses Conducted

1. Introduction to IoT
2. Object Oriented Programming
3. Computer Architecture and Organization using ARM

Amrita Robotics Research Center

Amrita Robotics Research Center (ARRL) was established in the year 2013. The center focuses on research and development in the field of robotics. The Department of Computer Science and Engineering has defined one of its research focus in the area of Robotics and its applications in societal benefits. The research is being carried out on the following topics:

• Emotion Recognition from Facial Expressions.
• Biped Navigation
• Effective Localization and Communication of Mobile Robots.
• Techniques for Mining Multimodal Multi-stream Data in a Robotic Environment.
• Mobile Robotics
• Industria Manipulators
• Swarm Robotics

The lab facilities include Bioloid- A humanoid robot kit, Lego Mindstroms robotic kit, Firebird-V from Nex Robotics with other sensors and equipments. Apart from this the center has few in-house built robots like AAMoR-I Service mobile robot platform, 5DoF dual Industrial Manipulator, a ingeniously built Humanoid Robot, Amrita Hexapod Robot, ADAM Manipulator

In-house Built Robots

Major Equipments Details

S. No	Equipment Name	Description/Specification	Quantity
1	Mobile Robotics platform- Fire Bird	ARM Based	5
2	Mobile Robotics platform- Fire Bird	ATMEGA Based	5
3	Bioloid premium Kit	Humanoid Robot Kit	1
4	LEGO Mindstroms Set	Lego Base Kit + Education Resource Set	2
5	sbRIO-9623	400 MHz Controller and LX25 FPGA, 12 Bit AIO, RMC Support	2
6	NI WSN Starter Kit	Kit includes a gateway, 2 nodes, batteries, sensors	1

Software Details

• Matlab
• LEGO MINDSTORMS Education NXT Software v.2.0
• LabVIEW Robotics Software Bundle

Key Projects

1. Facial Expression Generator of Emotions for Human-Robot Interaction
2. Emotion Synthesis from Speech
3. Stability and Control of a Biped Robot Walking on an Uneven Terrain
4. Humanoid Robot-Control and Stabilization for navigation on an uneven terrain
5. Navigation Techniques for Autonomous Mobile Service Robot