Imran Ahmed

Prathmesh Sawant, Yashwant Temburu, Mandar Datar et al.

Depth-map is the key computation in computer vision and robotics. One of the most popular approach is via computation of disparity-map of images obtained from Stereo Camera. Semi Global Matching (SGM) method is a popular choice for good accuracy with reasonable computation time. To use such compute-intensive algorithms for real-time applications such as for autonomous aerial vehicles, blind Aid, etc. acceleration using GPU, FPGA is necessary. In this paper, we show the design and implementation of a stereo-vision system, which is based on FPGA-implementation of More Global Matching(MGM). MGM is a variant of SGM. We use 4 paths but store a single cumulative cost value for a corresponding pixel. Our stereo-vision prototype uses Zedboard containing an ARM-based Zynq-SoC, ZED-stereo-camera / ELP stereo-camera / Intel RealSense D435i, and VGA for visualization. The power consumption attributed to the custom FPGA-based acceleration of disparity map computation required for depth-map is just 0.72 watt. The update rate of the disparity map is realistic 10.5 fps.

Omar Khan, Imran Ahmed, Joshua Cottingham et al.

The effectiveness of simple sensory cues for retraining gait have been demonstrated, yet the feasibility of humanoid avatars for entrainment have yet to be investigated. Here, we describe the development of a novel method of visually cued training, in the form of a virtual partner, and investigate its ability to provide movement guidance in the form of stepping. Real stepping movements were mapped onto an avatar using motion capture data. The trajectory of one of the avatar step cycles was then accelerated or decelerated by 15% to create a perturbation. Healthy participants were motion captured while instructed to step in time to the avatar's movements, as viewed through a virtual reality headset. Step onset times were used to measure the timing errors (asynchronies) between them. Participants completed either a visual-only condition, or auditory-visual with footstep sounds included. Participants' asynchronies exhibited slow drift in the Visual-Only condition, but became stable in the Auditory-Visual condition. Moreover, we observed a clear corrective response to the phase perturbation in both auditory-visual conditions. We conclude that an avatar's movements can be used to influence a person's own gait, but should include relevant auditory cues congruent with the movement to ensure a suitable accuracy is achieved.

Aniruddh Raghu, Matthieu Komorowski, Imran Ahmed et al.

Sepsis is a leading cause of mortality in intensive care units and costs hospitals billions annually. Treating a septic patient is highly challenging, because individual patients respond very differently to medical interventions and there is no universally agreed-upon treatment for sepsis. In this work, we propose an approach to deduce treatment policies for septic patients by using continuous state-space models and deep reinforcement learning. Our model learns clinically interpretable treatment policies, similar in important aspects to the treatment policies of physicians. The learned policies could be used to aid intensive care clinicians in medical decision making and improve the likelihood of patient survival.

Shubhra Aich, Anique Josuttes, Ilya Ovsyannikov et al.

In this paper, we investigate estimating emergence and biomass traits from color images and elevation maps of wheat field plots. We employ a state-of-the-art deconvolutional network for segmentation and convolutional architectures, with residual and Inception-like layers, to estimate traits via high dimensional nonlinear regression. Evaluation was performed on two different species of wheat, grown in field plots for an experimental plant breeding study. Our framework achieves satisfactory performance with mean and standard deviation of absolute difference of 1.05 and 1.40 counts for emergence and 1.45 and 2.05 for biomass estimation. Our results for counting wheat plants from field images are better than the accuracy reported for the similar, but arguably less difficult, task of counting leaves from indoor images of rosette plants. Our results for biomass estimation, even with a very small dataset, improve upon all previously proposed approaches in the literature.

M. Amine Chelihi, Akintunde Elutilo, Imran Ahmed et al.

Mobile phones have been playing a very significant role in our daily activities for the last decade. With the increase need for these devices, people are now more reliant on their smartphone applications for their daily tasks and many prefer to save their mobile data on a cloud platform to access them anywhere on any device. Cloud technology is the new way for better data storage, as it offers better security, more flexibility, and mobility. Many smartphones have been investigated as subjects, objects or tools of the crime. Many of these investigations include analysing data stored through cloud storage apps which contributes to importance of cloud apps forensics on mobile devices. In this paper, various cloud Android applications are analysed using the forensics tool XRY and a forensics taxonomy for investigation of these apps is suggested. The proposed taxonomy reflects residual artefacts retrievable from 31 different cloud applications. It is expected that the proposed taxonomy and the forensic findings in this paper will assist future forensic investigations involving cloud based storage applications.