Gopi Krishna Tummala

RoadMap: mapping vehicles to IP addresses using motion signatures

Inter-vehicular communication (IVC) can be used to enhance the sensing region of vehicles for improved safety on the roads. For many applications based on IVC, the relative locations and communication identities (e.g., IP addresses) of other collaborating vehicles are important for accurate identification. This is particularly challenging to achieve in the presence of legacy vehicles which may not have any sensing or IVC capabilities. We present a system called RoadMap, that matches IP addresses with respective vehicles observed through a camera. It assumes a smartphone or a dashboard camera deployed in vehicles, to identify the vehicles in field of view (FoV), and IVC capability. It runs in the adopted vehicles and accurately matches information obtained through multiple sensing modalities (e.g., visual and electronic). RoadMap matches the motion-trajectories of vehicles observed from the dash-board camera with the motion-trajectories transmitted by other vehicles. To the best of our knowledge, RoadMap is the first work to explore motion-trajectories of vehicles observed from a camera to create a map of vehicles by smartly fusing electronic and visual information. It has low hardware requirement and is designed to work in low adoption rate scenarios. Through real-world experiments and simulations, RoadMap matches IP-Addresses with camera observed vehicles with a median matching precision of 80%, which is 20% improvement compared to existing schemes.

Roadview: Live View of On-Road Vehicular Information

Inter-vehicular communication (IVC) can be explored to build a live map of vehicles for enhancing collaborative vehicular applications related to traffic statistics, safety by accident prediction and prevention, and energy efficient route planning. For enhancing the abovementioned applications, a live map of vehicles associated with their communication identities (e.g., IP/MAC addresses) of other collaborating vehicles is needed. This is particularly challenging to achieve in the presence of legacy vehicles which may not have any sensing or IVC capabilities. Additionally, vehicles might have diverse sensing capabilities and can have contradicting opinions of surrounding vehicles. We present RoadView, a system that builds the live map of surrounding vehicles by intelligently fusing the local maps created by individual vehicles. RoadView runs on top of existing local vehicular matching systems (LM) such as Foresight [10] or RoadMap [20]. RoadView is the first work that provides a live map of vehicles by leveraging collaboration across vehicles. Our simulations show that for different adoption rates and traffic densities, RoadView can robustly fuse information from a collection of local maps and enhance vehicles to sense 1.8x (average) number of immediate neighboring vehicles compared to state of art LM algorithms.

In-band wireless cut-through: is it possible?

This paper explores if wireless cut-through is possible. Unlike wired cut-through, wireless cut-through, if realized, can reduce latency and improve throughput by upto 3x. This paper shows that one way to implement cut-through is for every node to forward the previously decoded packet while receiving the current packet. This strategy is called decode-and-forward D&F). Another way is to have every node forward without decoding. This strategy is called amplify-and-forward (A&F). This paper shows that both D&F and A&F have many shortcomings. It proposes a new way to realize cut-through that reduces latency and increases throughput over traditional routing. This paper implements the first cut-through routing modules. Through multiple emulations of existing operational networks, it shows that cut-through switching can improve sum network throughput by up to 2.2x compared to traditional routing.

Soft-swipe: enabling high-accuracy pairing of vehicles to lanes using COTS technology

In this paper we demonstrate a novel system Soft-Swipe, which can enable highly accurate pairing of vehicles to respective lanes in a wide-range of vehicle-based multi-lane service stations using economical general-purpose commodity communication and sensing technology. To study the system, we consider an example application of pairing vehicles to respective quality check bays in an automobile manufacturing plant. Our proposed system called Soft-Swipe works by matching natural signatures (specifically, motion signatures) generated by the target object with the same signature detected by simple instrumentation of the environment (a video camera). Soft-Swipe is the first work that accurately captures the fine grain motion profile of vehicles using commodity hardware to provide vehicle-to-infrastructure pairing.

Vision-track: vision based indoor tracking in anchor-free regions

Smart-devices can render high quality location services when endowed with the ability to analyze information conveyed through video feed. In this paper, we aim to provide tracking services by using a mobile smart camera such as in google glasses and smartphones considering the following three objectives: (1) No additional deployment, (2) No user-side instrumentation or hardware upgrades, and (3) Easy adoption in practice. Existing RF or VLC based solutions for indoor tracking can provide location and orientation only when there are dense deployments of APs or VLC bulbs (anchor points) in users field of view. Vision-Track is the first vision based solution that can track the cameras location and orientation indoors even when no anchor point is in line-of-sight (LOS). Vision-Track deployed in an indoor college building provides a median localization accuracy of 49 cm.

Visualloc: vision based localization using a single smart-bulb

Smartphones can derive position and orientation by decoding the VLC messages from three or more lights. To achieve this goal, VLC based positioning solutions like Luxapose [3] require all the COTS luminaires to be modified. We propose VisualLoc, a system that enables smartphones to derive position and orientation using a composite-bulb which consists of a single modified commodity luminaire attached to a microcontroller and additional ordinary bulbs arranged in some geometric fashion. The modified luminaire flashes rapidly advertising its global coordinates encoded in terms of light intensity changes imperceptible to human eyes. The relative closeness and/or ordering of the bulbs w.r.t the identified smart-bulb can be leveraged to uniquely map the bulbs. A smartphone camera identifies the composite-bulb from a short video feed and decodes its location. Using this information, the global coordinates and orientation of the camera can be obtained with sub-meter level accuracy.

Autocalib: automatic traffic camera calibration at scale

Emerging smart cities are typically equipped with thousands of outdoor cameras. However, these cameras are typically not calibrated, i.e., information such as their precise mounting height and orientation is not available. Calibrating these cameras allows measurement of real-world distances from the video, thereby, enabling a wide range of novel applications such as identifying speeding vehicles, city road planning, etc. Unfortunately, robust camera calibration is a manual process today and is not scalable. In this paper, we propose AutoCalib, a system for scalable, automatic calibration of traffic cameras. AutoCalib exploits deep learning to extract selected key-point features from car images in the video and uses a novel filtering and aggregation algorithm to automatically produce a robust estimate of the camera calibration parameters from just hundreds of samples. We have implemented AutoCalib as a service on Azure that takes in a video segment and outputs the camera calibration parameters. Using video from real-world traffic cameras, we show that AutoCalib is able to estimate real-world distances with an error of less than 12%.

Navigation Assistance for Individuals with Visual Impairments in Indoor Environments

Canes or service dogs in indoor environments are unable to provide spatial information to the Individuals with Visual Impairments (IVIs) to make them independent. An indoor navigation assistance system can provide information on the presence of any obstacles in their vicinity, the distance of separation and their direction of motion (in case of mobile objects) w.r.t the IVIs. In this paper, we attempt to address the above objective by designing a novel time-efficient algorithm where a smart-glass is employed to spot an obstacle (stationary or mobile) in indoor environment using the inbuilt camera and inertial sensors. The system is implemented and tested extensively in indoor settings.