close× Call Us +1 (777) 123 45 67

Rıza Alp Güler, PhD

Imperial College London, Ariel AI

I am currently enrolled as a post-doctoral resarcher at Imperial College London. I hold a PhD degree in applied mathematics and informatics from INRIA and École Centrale Paris. I co-founded Ariel AI to bring my research ideas to products that will reach outside academia and enable the next generation of augmented reality experiences.

  • 2.6.2020:
    Our paper on bone-level parametric human body modellingn accepted at ECCV 2020 as a spotlight. Congrats Haoyang !
  • 10.4.2020:
    Our paper on in-the-wild hand reconstruction accepted at CVPR 2020 as an oral presentation. Congrats Dominik !
  • 8.3.2019:
    Two papers accepted at CVPR 2019 as oral presentations.
  • 27.11.2018:
    I receive STIC Doctoral School Best Sceintific Contribution Award 2018.
    “DensePose” has received the second place (ex aequo) in the annual award for PhD student research projects. ( link ) The doctoral school includes many renowned French universities and “grandes écoles”, including CentraleSupélec, Ecole Polytechnique, ENS Cachan, Télécom ParisTech, Université Paris-Sud, etc.
  • 25.7.2018:
    Two papers accepted at ECCV 2018.
    We co-organize COCO and PoseTrack challange workshops, introducing the new "DensePose" task.

Research Interests

My current research is mainly on neural networks for various visual tasks. I also work on sequential structured prediction models for dense labelling tasks. During my masters studies I worked on novel PDE based shape representation techniques and medical image analysis.

See Publications.


Coding Shape Inside the Shape: Regional Shape Representations
MSc Thesis
Sabanci University Faculty of Engineering and Natural Sciences Electronics Engineering Program

Supervised by: Dr. Gozde Unal

The shape of an object lies at the interface between vision and cognition, yet the field of statistical shape analysis is far from developing a general mathematical model to represent shapes that would allow computational descriptions to express some simple tasks that are carried out robustly and effortlessly by humans. In this thesis a novel perspective on shape characterization is presented: encoding shape information inside the shape. The representation is free from the dimensions of the shape, hence the model is readily extendable to any shape embedding dimensions (i.e 2D, 3D, 4D). A very desirable property is that the representation possesses the possibility to fuse shape information with other types of information available inside the shape domain, an example would be reflectance information from an optical camera.

Three novel fields are proposed within the scope of the thesis, namely `Scalable Fluctuating Distance Fields', `Screened Poisson Hyperfields', `Local Convexity Encoding Fields', which are smooth fields that are obtained by encoding desired shape information. `Scalable Fluctuating Distance Fields', that encode parts explicitly, is presented as an interactive tool for tumor protrusion segmentation and as an underlying representation for tumor follow-up analysis. Secondly, `Screened Poisson Hyper-Fields', provide a rich characterization of the shape that encodes global, local, interior and boundary interactions. Low-dimensional embeddings of the hyper-fields are employed to address problems of shape partitioning, 2D shape classification and 3D non-rigid shape retrieval. Moreover, the embeddings are used to translate the shape matching problem into an image matching problem, utilizing existing arsenal of image matching tools that could not be utilized in shape matching before. Finally, the `Local Convexity Encoding Fields' is formed by encoding information related to local symmetry and local convexity-concavity properties.

The representation performance of the shape fields is presented both qualitatively and quantitatively. The descriptors obtained using the regional encoding perspective outperform existing state-of-the-art shape retrieval methods over public benchmark databases, which is highly motivating for further study of regional-volumetric shape representations.

Undergraduate Thesis
Sabanci University Faculty of Engineering and Natural Sciences Electronics Engineering Program

Supervised by: Dr. Gozde Unal

Appeared in MICCAI Tractography Challenge Workshop

Diffusion Tensor Imaging provides micro-structural information about direction and restrictiveness of water diffusion in brain tissues. This makes DTI a great tool for analysis of the white matter anatomy, due to the restricted diffusion caused by the myalinated axon fibers. The can be reconstructed using a technique called 'tractography'. Clustering techniques for brain white matter voxels, which partition different regions of brain white matter tissue, are essential processing steps for quantitative analysis of DTI. Motivated by fusing the information from the tractography with the tensor clustering information, a novel iterative approach is described to extract the neaural tracts of interest from a diffusion tensor image.

Teaching Assistant Experience