Professors W. Heidrich P. Wonka M. Hadwiger P. Richtarik G. Sundaramoorthi B. Ghanem D. Michels
People Faculty Researchers Postdocs Students
Events Conference Seminar Workshop Sci-Café Social Events
Research Discovery Publications

FalconViz Gets The Upper Hand As Saudi Arabia’s Only Licensed Commercial Drone Operator

Jan 16, 2017

In the summer of 2014, people in Jeddah’s historical Al Balad neighborhood caught a curious sight. Unmanned aerial vehicles, commonly known as drones, buzzed 50 meters above the old city, prowling over its densely-packed stone rooftops.

The drones weren’t monitoring Al Balad for defense-related purposes—what unmanned aerial vehicles are typically known for. FalconViz sent them out. The startup, which operates out of King Abdullah University of Science and Technology (KAUST), developed a new way to survey and map areas, using high-resolution cameras mounted on drones to collect data.

After UNESCO designated Al Balad a World Heritage Site, Jeddah’s Municipality commissioned FalconViz to survey the district’s 250,000 square meters. The startup then ran the pictures through proprietary imaging software, transforming them into highly detailed 3D maps, plans and models. Historians and city planners used the renderings as part of ongoing preservation efforts.

FalconViz’s drones—which are no larger than a hobbyist’s remote control plane—can scan a 5 square kilometer urban area in little more than one week. In rural settings, they can complete the task in a few days.

“If you surveyed an area like that [Al Balad] with a traditional survey team it could take months,” says Neil Smith, co-founder and CEO of FalconViz.

The drones may be eye-catching, but the software sets FalconViz apart. It allows the company to take the data gathered by drones and turn it into life-like 3D maps in a few hours.

“We’re the only company who can provide [an] end-to-end service solution when it comes from data capturing, all the way to visualizing the data,” says Anas Dahlawi, FalconViz’s 33-year-old general manager. “We know we’re ahead of the game.”

By miles. So far, FalconViz is the only commercial drone operator with a government license.

Last year, Saudi Arabia’s General Authority of Civil Aviation made it illegal to operate drones without a permit. It singled out drones equipped with high-resolution cameras as a security threat, in addition to being a safety hazard.

It took more than a year and a chunk of money for Dahlawi to get a license in early 2016; he won’t disclose the amount. FalconViz’s four drone operators had to pass a commercial pilot’s test under rules that are stricter than those of the U.S. Federal Aviation Administration.

Although it has that advantage, FalconViz doesn’t have the surveying market to itself. It faces an established industry, where oil and gas companies have long required such services.

Competitors include Arabian Geophysical & Surveying Co. and Saudi Comet Co. Large construction companies, such as Al Jaber Group, also employ their own surveyors.

A traditional survey team working on the ground—using tripod mounted theodolites, levels, surveying poles and GPS—will methodically measure distances and angles between points by using triangulation. “This is a very time consuming process, requiring very skilled surveyors,” says Smith.

Satellites, airplanes, and helicopters can perform surveys, but aren’t always suitable for high resolution scans of smaller areas, like Al Balad. The surveys are not detailed enough. FalconViz’s technology provides resolution on a granular level—as high as half a centimeter.

Its software recognizes changes in camera positions between each image, and much like a traditional surveyor, uses that to triangulate the position of objects and points as they shift slightly from image to image.

The drones range from a spindly six-rotor copter designed for precision scanning, to an imposing fixed-wing drone that resembles a bat, and is used for long range flights.

FalconViz provides services for a variety of purposes, including monitoring construction sites, performing mining assessments, and urban planning. To date, it has completed 30 projects for 18 clients, such as Saudi Aramco, Aecom, and Dow Chemical Company. Saudi Binladin Group, which is expanding the area around Makkah’s Holy Mosque, recently signed a contract.

Its fees vary, but a survey covering a 5-kilometer urban area costs between $200,000 and $500,000, depending on the complexity of the project. When Jeddah’s Municipality enlisted FalconViz, a member told Smith it had recently spent $3.2 million on a land survey.

So far, he and Dahlawi have raised a total of $1.1 million from Saudi Aramco’s Wa’ed, which provides funding for entrepreneurs, and KAUST. They’ve amassed a fleet of eight drones and built a team of 18 employees.

It all started with one tiny drone and an impromptu experiment.

Neil Smith, a 38-year-old American anthropologist, arrived at KAUST in 2013 to conduct research at the university’s Visual Computing Center. He was drawn to KAUST, because Saudi Arabia, like the rest of the Middle East, features a wealth of historical sites in need of preservation, and he wanted to apply computational sciences, such as 3D reconstruction, to archaeology.

At KAUST, Smith kept hearing about Luca Passone, a PhD candidate in earth sciences. Passone, 31, was experimenting with drones, building them from scratch by importing the frames, motor components, batteries, remote control transmitters and receivers, and adding cameras. A fellow researcher suggested Smith could perhaps use Passone’s drones for his imaging work.

Smith and Passone hit it off, and decided to test whether the drones could capture the data Smith needed to create accurate 3D models and maps.

Passone went to the campus mosque one morning in late 2013. Standing on the lawn under the minaret, he directed the drone over the tower, its cameras trained downwards. He then sent the images to Smith, who processed them using modeling software he developed. “Within a couple of hours we had the entire mosque captured in 3D,” says Smith. “We were blown away.”

They invited Mohamed Shalaby, a 48-year-old mathematician at the Visual Computing Center, to help them with geometric modeling.

In early 2014, a group of KAUST industry partners touring the campus stopped to view a demo of FalconViz’s technology. Afterwards, a representative from Jeddah’s Municipality invited the trio over; the meeting landed them the Al Balad contract. “I think that’s when we first realized that this could be used heavily for commercial applications,” says Smith.

He and his partners used a $205,000 grant from KAUST to build a new drone and fine-tune their software, but they needed business help. “We come from scientific backgrounds,” says Shalaby.

They turned to Dahlawi, a frequent visitor at the Visual Computing Center. He agreed to temporarily manage the business. Dahlawi has a master’s degree in project management from George Washington University School of Business, and is a founder of a sports marketing company.

He knew how to navigate the Saudi business world—at least more than three foreign academics. Six months later, Smith and his partners asked Dahlawi to stay full-time. “We grafted him into the founders,” says Smith. Dahlawi was responsible for obtaining the drone license, allowing FalconViz to continue operations in Saudi Arabia.

From there the startup took flight. The Al Balad project led to another contract from the Jeddah Municipality to map out a seaside boulevard in 2015. FalconViz also scored a contract with Makkah’s Municipality.

For many of its clients, such as Saudi Aramco, it has to submit to confidentiality agreements, but other projects are innocent enough. In 2015, FalconViz partnered with the American University of Beirut to survey the iconic Pigeon Rocks, a natural stone arch formation rising from the Mediterranean just off the city’s corniche. The university used the data as part of a virtual reality experiment.

But, FalconViz doesn’t have immediate plans to spread beyond Saudi Arabia. Most countries in the region have either strict or ambiguous regulations around drones. “Here, we can prove the technology that we’re creating really is transformational,” says Smith.

Forbes Middle East rankings for VCC startups

Dec 15, 2016

Three startups of VCC have been ranked among the top 50 startups in Saudi Arabia by Forbes Middle East. The startups are FalconViz (5), Visual Experience (19), and Hachid (21).

Forbes Middle East elaborates:
"Earlier this year, Saudi Arabia unveiled Vision 2030, its plan to diversify away from oil. Important players in that shift will be entrepreneurs.

Although some big investors showed no sign of activity in 2016, notably the corporate venture arms of Saudi Telecom and Mobily, others, such as Saudi Aramco’s Wa’ed and Qotuf AlRiyadah/Flat6Labs Jeddah, are spending time and money developing a startup ecosystem. Tuba Terekli, who heads Qotuf AlRiyadah/Flat6Labs Jeddah, has invested in 38 companies so far, including mobile game developer Ayah Studios, which has nearly two million downloads, and Maharah, a provider of on-demand home maintenance services, which now employs 450 people.

Startups are also getting a boost from universities. King Abdullah University of Science and Technology (KAUST) runs KAUST Innovation Fund that has made investments in NOMADD, a manufacturer of automated solar panel cleaning technology, and Sadeem, which makes sensors. King Fahd University of Petroleum and Minerals' Dhahran Techno Valley also incubates startups and provides seed investment; they include Telaa, which developed anti-corrosion material for pipelines that is supposed to be more effective and environmentally friendly than existing technology, and ConCure which uses less water in making concrete.

We narrowed the list this year to 50, down from 100, and came up with stronger candidates. Twenty-four are new, including C3 Films, Cura and Rwaq. E-commerce still dominates, but startups are becoming savvier at using technology to collect and analyze data from transactions. Examples include OrderMe and Morni. Out of the 50 startups, 16 are still bootstrapping.

Methodology: We looked at more than 250 startups, and ranked them according to the amount of equity raised (startups whose sole funding came from loans were excluded), and source of funding. We also took into consideration business models and the potential size of the market. The majority of entrepreneurs were forthcoming and transparent about their funding, but a few asked us to publicly withhold that information. Startups are maximum six years old."

Technical Papers Chair - Wolfgang Heidrich

Oct 05, 2016

​Preview the vision of SIGGRAPH Asia 2016 Technical Papers from Wolfgang Heidrich, SIGGRAPH Asia 2016 Technical Papers Chair.
See Video

CS Graduate Seminar | On solving large-scale MRFs on GPUs and on evaluating image-based modeling and rendering systems

Sep 26, 2016

By Professor Michael Goesele ( TU Darmstadt, Germany )

Image-based modeling and rendering systems have made tremendous progress in recent years. Motivated by our own modeling and rendering pipeline, I will present two separate contributions:First, I will discuss the challenges in evaluating the end-to-end quality of image-based modeling and rendering system. I will then introduce our proposed methodology, discuss several quality metrics and show its relationship to existing benchmarks. One of the key advantages of this approach is that it does not require ground truth geometry. Finally, I will conclude the first part of my talk by introducing our new image-based modeling and rendering benchmark. In the second part of the talk, I will focus on solving large-scale MRFs, one of the key bottlenecks in the texturing stage of our reconstruction pipeline. I will first discuss strategies to parallelize the solution of large scale MRFs with a focus on massively-parallel approaches suitable for modern GPUs. I will then introduce our solver that supports arbitrary MRF topologies efficiently and can handle arbitrary, dense or sparse label sets as well as label cost functions. Together with two additional heuristics for further acceleration, our solver performs favorably even compared to modern specialized solvers in terms of speed and solution quality, especially when solving very large MRFs. This is joint work with Michael Waechter, Mate Beljan, Simon Fuhrmann, Nils Moehrle, Johannes Kopf, Daniel Thuerck, Sven Widmer, Max von Buelow, Patrick Seemann and Marc E. Pfetsch.

Biography: Dr.-Ing. Michael Goesele is a professor in TU Darmstadt’s Department of Computer Science where he heads the research group “Graphics, Capture and Massively Parallel Computing”. He received his diploma in Computer Science from Ulm University in 1999. He then joined the Max Planck Institute for Computer Science in Saarbrücken and earned his Ph.D. from Saarland University in 2004. After a two year postdoctoral stay at the University of Washington (Seattle), he joined TU Darmstadt in 2007. His research interests include computer graphics, computer vision, and massively parallel computing.

Peter Wonka's new query tool, called a relation-augmented image descriptor will change forever querying images online.

Aug 18, 2016

Today, querying the massive amounts of images available in online databases such as Flickr and Instagram can be a time-consuming experience and sometimes even ineffective. Researchers from the King Abdullah University of Science and Technology (KAUST), Saudi Arabia, and the University College London, have developed a new tool that generates image queries based on a geometric description of objects in spatial relationships with potential applications in computer graphics, computer vision, and automated object classification.

The collection of photographs and pictures available on the internet represents an immense source of data that can be “read” and used only if we use queries capable of sorting out images according to their salient attributes.
"At present, when searching for images on the internet, the query is based entirely on text descriptions that go along with them," says Peter Wonka, the researcher from KAUST who designed and led the study. "For being effective a text description needs to be short and to provide true and pertinent information on the picture it represents."
Wonka and his colleagues Paul Guerrero and Niloy Mitra from University College London decided to add something more powerful to the text-based search tool currently available.
"The obstacle was defining an extra descriptor without adding extra metadata to existing images”, explained Wonka “ and we went for indicators of existing relationships between objects—such as 'riding,' 'carrying,' 'holding' or 'standing on'. In other words, we added verbs to the nouns describing the images.”
The new query tool called a relation-augmented image descriptor (RAID), is able to analyze the textual description while searching for matches using geometric analogies.
"RAID allows queries using complete textual descriptions such as 'a person standing on snowboard' as well as images or simple sketches of the desired composition of objects," said Wonka. "The new tool uses spatial descriptors such as 'above' or 'left of' – which allow discriminating among different complex relationships."
The team is currently working on a three-dimensional version of the descriptor able to search for extra details in the pictures.

More Information

KAUST SG paper 'Sketch-Based Query for Searching for Relationships Among Objects in Images' was cited by ACM tech news

Aug 11, 2016

​Researchers at the King Abdullah University of Science and Technology (KAUST) and the University College London have proposed a tool that generates image queries based on a sketch or description of objects in spatial relationships, which could make it easier to search large databases of images.

Deep Learning Summer School

Aug 01, 2016

I participated in the Deep Learning Summer School organized by the CIFAR Institute and the CRM Center of Canada. This event took place at the University of Montreal, and it consisted of a series of lectures covering a broad range of topics about deep learning from a machine learning perspective touching multiple applications like computer vision, natural language processing, and robotics. The talks were given by top researchers in each area working in academia (NYU, MIT, Stanford) and industry (Google, Facebook, Twitter, etc.) such as Rob Fergus, Jeff Dean, Yousha Bengio among others.
It was the second edition of this summer school, which it is pretty selective, with an acceptance rate around 25%. In that sense, it was a pleasant experience for me to talk with highly competitive people from academia and industry working at the heart of deep learning models. It was a really insightful experience to hear about the way other researchers are tackling their specific problems as well as their perception of the field and the relevant challenges in the next couple of years. Another essential part of the summer school was the opportunity to present recent and ongoing research work, I took advantage of it to talk and discuss my latest work accepted for ECCV-2016. Our work was received well from practitioners and researchers, a group of people is eager to try it out for specific applications such as sports analytics and pedestrian analysis, and others are interested in built on top of our observations.
In summary, it was a splendid time to learn, network with people and spread the presence of KAUST through the work of a member of the Visual Computing Center.

Hachid Receives Award At View / CVPR 2016

Jul 20, 2016

Hachid won the award for the Venture pitch with the best business plan at the Fifth Annual Vision Industry and Entrepreneur Workshop (VIEW) 2016. VIEW is held in conjunction with the prestigeous and highly competitive Conference on Vision and Pattern Recognition, one of the premier scientific venues in the field of computer vision. VIEW brings together young high-tech startups and established industry in the field of Computer Vision, Robotics, and Graphics.

We at hachid feel deeply honoured by this award, which we received in competition with three other cutting edge high-tech startups.

My Remarkable Summer School Experiences

Jul 17, 2016

My extraordinary journey started at the eleventh Microsoft Research Summer School, which took place in Cambridge, UK. I had the opportunity to attend a series of talks of academic interest and transferable skills conducted by world leader researchers from Microsoft. Additionally, I was able to present my research during a poster session where I had the chance to meet several students and Microsoft researchers working on related research topics. Overall, it was a remarkable summer school that balanced between professional development and social activities.
The next experience included my visit to the fabulous city of Sicily, Italy, where the International Computer Vision Summer School (ICVSS) 2016 was hosted. This summer school is one of the most renowned in Computer Vision with a selection process that included only 30% of the total student candidates to attend the event. The tenth edition of the ICVSS provided me an objective and clear overview of the state-of-the-art research in Computer Vision. At the same time, I had a golden opportunity to meet and learn about internship opportunities at world leading companies such as Google, Facebook, and Microsoft. Doubtless, ICVSS is a must to attend for all Computer Vision Ph.D. Students.

VCC Spin-off Patents Technology

Jul 06, 2016

Hachid Arabia, a VCC spin-off startup, has been granted an international and US patent entitled “System and Method for Crowd Counting and Tracking” under Application Nos. The patent secures intellectual property to an embedded computing platform and proprietary algorithm to analyze sensor data in order to perform head counting in confined spaces.

Highest paper approvals for CVPR

Jun 01, 2015

Congratulations to all faculty and students for this great achievement.

Computer vision: Teaching computers how to see the world

May 27, 2015

Ghanem works in the Visual Computing Center (VCC) and is interested in computer vision, image processing and machine learning. "People in computer vision are trying to teach machines how to see the world the same way we see the world. Seeing not only object color and so forth, but actually understanding semantically what this visual data is all about," he described.​

And The Winners are...

Mar 08, 2015

Congratulations to all the participants thank you for all your hard work for this poster competition.​

NeuroLines: A Subway Map Metaphor for Visualizing Nanoscale Neuronal Connectivity [INFOVIS HONORABLE MENTION AWARD]

Oct 19, 2014

NeuroLines: A Subway Map Metaphor for Visualizing Nanoscale Neuronal Connectivity
Authors: Ali Al-Awami, Johanna Beyer, Hendrik Strobelt, Narayanan Kasthuri, Jeff W. Lichtman, Hanspeter Pfister, Markus Hadwiger

Visual Computing Center | Summit 2014

Mar 26, 2014

KAUST VCC’s Summit is an exciting opportunity to get together with visual computing experts from KAUST and abroad.
The workshop will feature a series of invited talks by Professors, researchers and Industry representatives from all over the world.
Hosted by Visual Computing Center, the venue for the summit will be at KAUST campus, located on the coast of the Red Sea in Thuwal, Kingdom of Saudi Arabia.