[2018-09-15] Attended ECCV 2018 in Munich, Germany and presented our two accepted papers.
[2018-07-03] 2 papers (Action Search and DETAD) accepted to ECCV 2018!!
[2018-07-14] Attended ICVSS 2018 summer school in Sicily.
[2018-06-22] Presented our DETAD work in the ActivityNet challenge workshop in CVPR18 (slides) and also released the code for the DETAD diagnosis tool.
[2018-06-01] I’ll attend CVPR18. Come check out our ActivityNet challenge workshop on Friday, June 22.
[2018-05-01] Started my PhD studies with Bernard Ghanem at KAUST. I’m continuing in the same research direction of video understanding and computer vision in general.
[2018-04-16] Got accepted to the ICVSS 2018 summer school in Sicily.
[2018-04-10] Successfully defended my Master’s thesis.
[2018-03-23] I’m co-organizing the third annual ActivityNet challenge in CVPR18, Salt Lake City (the challenge starts today). Check out our website. This year we have six exciting tasks and five novel action datasets.
[2017-12-15] Graduated with an MSc in Computer Science from KAUST.
[2017-05-01] I’m co-organizing the second annual ActivityNet challenge in CVPR17, Hawaii (the challenge starts today).
[2016-09-21] Started my Master’s degree in Computer Science at KAUST. I joined the multicultural and diverse Image and Video Understanding Lab (IVUL) advised by Bernard Ghanem.
[2016-06-06] Started a software development engineer internship at Amazon Corporate LLC, Seattle, WA with the Vendor Self Service, Business Advisor team.
[2016-05-29] Graduated from Cornell University with a Bachelors degree in both Computer Science and Mathematics.

Selected Publications

State-of-the-art temporal action detectors inefficiently search the entire video for specific actions. Despite the encouraging progress these methods achieve, it is crucial to design automated approaches that only explore parts of the video which are the most relevant to the actions being searched for. To address this need, we propose the new problem of action spotting in video, which we define as finding a specific action in a video while observing a small portion of that video. Inspired by the observation that humans are extremely efficient and accurate in spotting and finding action instances in video, we propose Action Search, a novel Recurrent Neural Network approach that mimics the way humans spot actions. Moreover, to address the absence of data recording the behavior of human annotators, we put forward the Human Searches dataset, which compiles the search sequences employed by human annotators spotting actions in the AVA and THUMOS14 datasets. We consider temporal action localization as an application of the action spotting problem. Experiments on the THUMOS14 dataset reveal that our model is not only able to explore the video efficiently (observing on average 17.3% of the video) but it also accurately finds human activities with 30.8% mAP.
In ECCV, 2018

Despite the recent progress in video understanding and the continuous rate of improvement in temporal action localization throughout the years, it is still unclear how far (or close?) we are to solving the problem. To this end, we introduce a new diagnostic tool to analyze the performance of temporal action detectors in videos and compare different methods beyond a single scalar metric. We exemplify the use of our tool by analyzing the performance of the top rewarded entries in the latest ActivityNet action localization challenge. Our analysis shows that the most impactful areas to work on are: strategies to better handle temporal context around the instances, improving the robustness w.r.t. the instance absolute and relative size, and strategies to reduce the localization errors. Moreover, our experimental analysis finds the lack of agreement among annotator is not a major roadblock to attain progress in the field. Our diagnostic tool is publicly available to keep fueling the minds of other researchers with additional insights about their algorithms.
In ECCV, 2018

Selected Projects

FPEAnalysis: Automating Floating Point Error Analysis [2016]

FPEAnalysis software system automates floating-point error analysis of scientific computing code using the “1+ δ” error model. FPEAnalysis analyzes a piece of code, automatically reports any numerically unstable subexpressions, and gives insights for resolving these instabilities.

Gates App: A Central SDN Controller for Gates Hall Network [2015]

Gates App is a software-defined networking controller for the 28-switch network of Gates Hall, home of Cornell’s computer science department. Gates App employs the Frenetic Project libraries to perform Ethernet routing and host discovery for a dynamically changing network.

ParEx: A Parallel Extrapolation Solver for Initial-Value ODEs [2015]

ParEx is an efficient shared-memory parallel implementation of explicit extrapolation methods for solving initial-value ordinary differential equations. Using a novel optimal load-balancing algorithm with adaptive step size and order control, ParEx outperforms state-of-the-art solvers on a range of test problems where the derivative evaluation is moderately expensive.

Professional Experience

The ActivityNet Large Scale Activity Recognition Challenge [2018]

Co-organizer and Program Chair
This challenge is the 3rd annual installment of the ActivityNet Large-Scale Activity Recognition Challenge, which was first hosted during CVPR 2016. It focuses on the recognition of daily life, high-level, goal-oriented activities from user-generated videos as those found in internet video portals. It was a full-day workshop held in CVPR 2018 in Salt Lake, Utah and held six diverse challenge tasks. It attracted a large number of participants from across the world, and it was sponsored by several industrial partners including Google DeepMind, Facebook, and Google AI.

Mantis Company [2017-Present]

Co-founder and Computer Vision Researcher
A state-of- the-art activity-based, advertising-centric automated video understanding platform. Mantis utilizes faster-than- real-time activity and object detection techniques for a fine-grained video content categorization to achieve a content-aware ads placement on videos.

The ActivityNet Large Scale Activity Recognition Challenge [2017]

Co-organizer and Program Chair
The ActivityNet Challenge was a half-day workshop held in CVPR 2017 in Honolulu, Hawaii. The challenge attracted a large number of participants from across the world, and it was sponsored by several industrial partners including Google DeepMind, NVidia, Qualcomm, and Panasonic.

Amazon Corporate LLC, Seattle, WA [2016]

Software Development Engineer Intern
Team: Vendor Self Service, Business Advisor. Manager: Ram Yerramilli.


KAUST Fellowship for MS and PhD Studies [2016-Present]

A fellowship which supports students for the duration of their graduate studies at KAUST. It includes ​full tuition support, monthly living allowance, housing, and medical coverage.

SACM Undergraduate Scholarship [2010-2016]

A scholarship awarded by the Saudi Arabian Cultural Mission to the United States. It covers the full tuition for an undergraduate STEM degree at a US university.

KAUST Gifted Student Program (KGSP) Scholarship [2010-2016]

KGSP is a prestigious scholarship awarded by KAUST to a select group of Saudi students, allowing them to pursue undergraduate degrees in STEM fields in the US, and then complete their master’s degree at KAUST.

Administrative Experience

Graduate Tutor [2018]

One-on-one tutoring sessions for graduate students in computer science and mathematics courses at KAUST.

Graduate Teaching Assistant  [2017]

CS240: Computing Systems and Concurrency, Professor Marco Canini, Computer Science, KAUST.  

Undergraduate Teaching Assistant  [2013]

Math Explorers’ Club, Mathematics Department, Cornell University
The Math Explorers’ Club is an NSF-supported project that develops materials and activities to give middle school and high school students an experience of more advanced topics in mathematics.


  • +966 551708480
  • humam-w
  • 4700 KAUST, Al Khawarizmi Building (Bldg 1), Office #2106-WS10, Thuwal 23955-6900, Saudi Arabia