anas@fanar:~$ cat about.md

# About

In my day-to-day business I am somewhere between research code and production
data pipelines — mostly enjoying Software Engineering as a craft while it
has not been killed yet through our unwavering quest to make ourselves
Skynet's prey! Prior to that I was concerned proving to my advisor I
deserve a PhD through working on 3D Computer Vision.

## what I work on

- MLOps for Fanar — building MLOps solutions for Fanar Arabic GenAI,
  addressing Effective Data Management, Data Analytics, Data-Training-
  Deployment Pipelines, and production-grade code.
- (3D) Computer Vision — worked on the alignment problem of
  partially-overlapping 3D point clouds during my PhD.
- Humanitarian AI — Computer Vision for remote-sensing data for
  disaster analysis.

## philosophy

> Only reviewed code is good code.

I like clear terminology, people who look at code as the outcome of
Software Engineering — not "coding" or "programming" — and people who
don't consider science a source for fame but rather a mission to deliver
societal impact.

anas@fanar:~$ ./stack.sh

[+] resolving toolchain...
[+] 5 layers loaded · 0 errors · YMMV

== Programming Languages ==  // as long as it still matters
   Python  ·  Scala  ·  C++

   // also did C# once upon a time, but Claude beats me to it 100×  ;(
   // there was a time when I considered Matlab a programming language
   //   (feeling ashamed). It is nonetheless a super powerful engineering tool.

== Orchestration ==
   pipeline frameworks   : Airflow · Kubeflow · AWS Step Functions
   containers / cluster  : Kubernetes · SLURM

== Databases ==
   PostgreSQL · SQLite · Elasticsearch · OpenSearch · AWS RDS / DynamoDB

== DevOps ==
   CI / CD     : Zuul CI · Jenkins
   deployment  : Docker · Ansible

== Big Data ==
   Spark · HDFS · MapR

[✓] stack initialized. mount points ready.

anas@fanar:~/publications$ ls -la | sort -k citations -r

citations : 487 h-index : 10 i10-index : 10 // canonical source: scholar.google.com/citations?user=APUSJ6UAAAAJ

JOURNAL

2011Mobile visual location recognition★ 153

G Schroth, R Huitl, D Chen, M Abu-Alqumsan, A Al-Nuaimi, E Steinbach

IEEE Signal Processing Magazine

Phone snaps a photo → image-retrieval against a geo-tagged database returns your location. GPS-free urban localization from pixels alone.

CONFERENCE

—Point Cloud-based Model-mediated Teleoperation with Dynamic and Perception-based Model Updating★ 89

X Xu, B Cizmeci, A Al-Nuaimi, E Steinbach

—

Stream a 3D point-cloud model of a remote scene to a haptic operator. Perception updates the model on the fly so teleop stays low-latency.

JOURNAL

2016Geometric-Algebra LMS Adaptive Filter and its Application to Rotation Estimation★ 41

WB Lopes, A Al-Nuaimi, CG Lopes

IEEE Signal Processing Letters

Reformulates the classical LMS adaptive filter in Clifford / geometric algebra — a clean online way to estimate 3D rotations from streaming data.

CONFERENCE

2016Analyzing LiDAR scan skewing and its impact on scan matching★ 28

A Al-Nuaimi, W Lopes, P Zeller, A Garcea, C Lopes, E Steinbach

International Conference on Indoor Positioning and Indoor Navigation (IPIN)

A rotating LiDAR distorts its own scan whenever the platform is moving. The paper quantifies the effect and compensates for it during scan matching.

CONFERENCE

20166DOF point cloud alignment using geometric algebra-based adaptive filtering★ 28

A Al-Nuaimi, E Steinbach, WB Lopes, CG Lopes

IEEE Winter Conference on Applications of Computer Vision (WACV)

Drives the geometric-algebra LMS filter to estimate full 6-DOF rigid alignment (rotation + translation) between two overlapping point clouds.

WORKSHOP

2013Performance comparison of various feature detector-descriptor combinations for content-based image retrieval with JPEG-encoded query images★ 28

J Chao, A Al-Nuaimi, G Schroth, E Steinbach

IEEE 15th International Workshop on Multimedia Signal Processing

Benchmarks SIFT / SURF / ORB / BRISK / FREAK on heavily JPEG-compressed queries — what survives the bandwidth squeeze for mobile image retrieval?

PATENT

2017Visual localization method★ 27

E Steinbach, G Schroth, M Abu-Alqumsan, R Huitl, A Al-Nuaimi

US Patent 9,641,981

US patent covering the end-to-end mobile visual-place-recognition pipeline: device-side query, server-side index, location response.

CONFERENCE

2011Rapid image retrieval for mobile location recognition★ 21

G Schroth, A Al-Nuaimi, R Huitl, F Schweiger, E Steinbach

IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)

Real-time variant of the localization pipeline — vocabulary-tree quantization + geometric verification fast enough to run on a phone.

JOURNAL

2012Fully automatic and frame-accurate video synchronization using bitrate sequences★ 11

F Schweiger, G Schroth, M Eichhorn, A Al-Nuaimi, B Cizmeci, M Fahrmair

IEEE Transactions on Multimedia

Syncs multiple video streams frame-accurately by cross-correlating their per-frame bitrate signatures. No audio fingerprint required.

WORKSHOP

2012Seamless and efficient stream switching of multi-perspective videos★ 10

H Zhang, A Al-Nuaimi, X Gu, M Fahrmair, R Ishibashi

International Packet Video Workshop

Streams multiple camera angles simultaneously and switches viewpoint without rebuffering — useful for live sports, concerts, 360° playback.

DISSERTATION

2017Methods of Point Cloud Alignment with Applications to 3D Indoor Mapping and Localization★ 7

A Al-Nuaimi

PhD Dissertation · Technische Universität München

The full PhD thesis: methods for aligning 3D point clouds, applied to indoor mapping and localization at building scale.

WORKSHOP

2015Indoor location retrieval using shape matching of KinectFusion scans to large-scale indoor point clouds★ 7

A Al-Nuaimi, M Piccolrovazzi, S Gedikli, E Steinbach, G Schroth

Eurographics Workshop on 3D Object Retrieval

Localize a small KinectFusion scan inside a much larger pre-built 3D map of a building, using shape descriptors and geometric matching.

CONFERENCE

2013Towards the design of an intuitive multi-view video navigation interface based on spatial information★ 6

S Apostu, A Al-Nuaimi, E Steinbach, M Fahrmair, X Song, A Möller

Mobile HCI · ACM

UI study for navigating multi-view video using the spatial relationship between camera views — context-aware switching that respects how viewers think about the scene.

CONFERENCE

2016Enhanced Depth Estimation using a Combination of Structured Light Sensing and Stereo Reconstruction★ 6

A Wittmann, A Al-Nuaimi, E Steinbach, G Schroth

VISAPP · Intl. Conf. on Computer Vision Theory and Applications

Combines a low-res structured-light prior with stereo disparity estimation in an α-expansion loop — denser depth maps, smaller search ranges, faster runtimes.

WORKSHOP

2013Towards location recognition using range images★ 6

A Al-Nuaimi, R Huitl, S Taifour, S Sarin, X Song, YX Gu, E Steinbach, M Fahrmair

IEEE ICME Workshops (ICMEW)

Adapts content-based image retrieval to range images for indoor localization. Benchmarks NARF against appearance features on a challenging indoor dataset.

WORKSHOP

2012ConCor+: Robust and confident video synchronization using consensus-based cross-correlation★ 5

A Al-Nuaimi, B Cizmeci, F Schweiger, R Katz, S Taifour, E Steinbach

IEEE 14th Intl. Workshop on Multimedia Signal Processing (MMSP)

Extends the bitrate-based video sync with a consensus-driven cross-correlation that survives dropouts and noise — and emits a confidence score, so you can reject untrustworthy alignments instead of silently shipping them.

PATENT

2019Visual localization method★ 5

M Abu-Alqumsan, A Al-Nuaimi, R Huitl, G Schroth, F Schweiger

US Patent 10,229,136

Continuation in the visual localization patent family — refines and extends the claim set around the mobile place-recognition pipeline first granted as US 9,641,981.

JOURNAL

20176DOF decoupled roto-translation alignment of large-scale indoor point clouds★ 4

A Al-Nuaimi, S Hilsenbeck, A Garcea, E Steinbach

Computer Vision and Image Understanding (CVIU) · Elsevier

A scalable alignment method for large indoor point clouds: decouple rotation from translation, solve each in a smaller search space, then refine jointly. Faster and more robust than coupled 6-DOF search.

PATENT

2018Visual localization method★ 3

M Abu-Alqumsan, A Al-Nuaimi, R Huitl, G Schroth, F Schweiger

US Patent 9,946,733

Continuation in the visual localization patent family — carves out a distinct claim scope from US 9,641,981.

PATENT

2023Visual localization method★ 1

M Abu-Alqumsan, A Al-Nuaimi, R Huitl, G Schroth, F Schweiger

US Patent 11,803,586

Latest continuation in the visual localization patent family — extends the original 2017 disclosure into a fresh claim set granted six years later.

anas@fanar:~$ git log --oneline --all --repos

[!] disclaimer: most of the code I write lives behind corporate firewalls
    and is highly protected — there is little public code to show here.
    The list below is what GitHub exposes; the real work is elsewhere.

loading...

anas@fanar:~$ curl -s https://medium.com/@anas.alnuaimi/feed | xmllint --xpath '//item' -

loading...

anas@fanar:~$ ./contact.sh

[+] resolving channels...
[+] handshake established. you have 4 vectors:

    ┌─ email     → analnuaimi@hbku.edu.qa
    ├─ github    → @TUMAAA
    ├─ linkedin  → /in/anasaa
    └─ medium    → @anas.alnuaimi

[✓] ready for incoming packets.