Dr. David Engel


Picture of Engel, David, Dr.

David Engel

Position: Research Scientist  Unit: Alumni Bülthoff


Computer vision aims to teach machines and algorithms to `see' with the ultimate goal of creating `intelligent' applications and devices that can provide assistance to humans in a wide array of scenarios. My approach investigates computer vision on three layers: low-level features, mid-level representations and high-level applications. Each of the layers depends on the previous ones while also generating constraints and requirements for them. At the application layer human-machine interfaces come into play and link the human perception to computer vision. By studying all layers we can gain a much deeper insight into the interplay of different methods, than by examining an isolated problem. Furthermore, we are able to factor constraints imposed by different layers and the users into the design of the algorithms, instead of optimizing a single method based purely on algorithmic performance measures. The different modules of my thesis are tightly connected and inter-dependent, in the framework of shape-centered representations. The connections between the modules avails the possibility to feed information back from higher to lower layers and optimize the design choices there. My work on these three layers includes:


Shape-Centered Features:

These interest points are formed at location of high local symmetry as opposed to corner interest points which occur along the outline of shapes. Experiments show that they are very robust with respect to common natural image transformations, such as scaling, rotation and the introduction of noise and clutter.


Mid-Level Representations:

I presented two strategies to build robust mid-level image representations:  First, a novel feature grouping method is introduced. The scheme offers a powerful way to combine the advantages of shape-centered interest points, namely robustness and a tight connection to a unique shape, and corner-based interest points, namely strong descriptors. Secondly, I introduced a novel set of medial feature superpixels. They represent a feed-forward way to divide the image into small, visually-homogeneous regions offering a compact and efficient mid-level representation of the image information.



Here, I bridge the gap between computer vision and the human observer by introducing three applications that employ the shape-centered representations from the two previous layers. The first step is a multi-class scene labeling scheme that produces dense annotations of images, combining a local prediction step with a global optimization scheme. I developed a novel image retrieval tool that operates on high-level semantic information allowing a more efficient image search in large labeled datasets. Finally, I put forward and investigated the novel idea of predicting the detectability of a pedestrian in a driver assistance context.


Further Projects:

I am currently working on open question in the fields of efficient crowd-sourcing and active learning. Here, the human user enters directly into the loop via crowd-sourcing services such as Amazon Mechanical Turk. This approach can allow us to efficiently optimize our tools as well as distribute complex tasks between human and machine intelligence (leaving only few hard examples for the human while the machine learning takes care of the vastly larger easy parts of a problem).

References per page: Year: Medium:

Show abstracts

Books (1):

Engel D: Shape-Centered Representations: From Features to Applications, 173, Logos-Verlag, Berlin, Germany, (2011). ISBN: 978-3-8325-2820-1, Series: MPI Series in Biological Cybernetics ; 27

Articles (2):

Engel D and Curio C (October-2013) Detectability Prediction for Increased Scene Awareness IEEE Intelligent Transportation Systems Magazine 5(4) 146-157.
Neth CT, Souman JL, Engel D, Kloos U, Bülthoff HH and Mohler BJ (July-2012) Velocity-Dependent Dynamic Curvature Gain for Redirected Walking IEEE Transactions on Visualization and Computer Graphics 18(7) 1041-1052.

Conference papers (8):

Engel D and Curio C (June-2012) Detectability prediction in dynamic scenes for enhanced environment perception, IEEE Intelligent Vehicles Symposium (IV 2012), IEEE, Piscataway, NJ, USA, 178-183.
Engel D, Herdtweck C, Browatzki B and Curio C (September-2011) Image Retrieval with Semantic Sketches In: Human-Computer Interaction: INTERACT 2011, , 13th IFIP TC13 Conference on Human-Computer Interaction, Springer, Berlin, Germany, 412-425, Series: Lecture Notes in Computer Science ; 6946.
Engel D and Curio C (June-2011) Pedestrian Detectability: Predicting Human Perception Performance with Machine Vision, IEEE Intelligent Vehicles Symposium (IV 2011), IEEE, Piscataway, NJ, USA, 429-435.
Neth C, Souman JL, Engel D, Kloos U, Bülthoff HH and Mohler BJ (March-2011) Velocity-Dependent Dynamic Curvature Gain for Redirected Walking, IEEE Virtual Reality Conference (VR 2011), IEEE, Piscataway, NJ, USA, 151-158.
Engel D and Curio C (June-2010) Shape Centered Interest Points for Feature Grouping, CVPR 2010 Workshop on Perceptual Organization in Computer Vision (POCV 2010), IEEE, Piscataway, NJ, USA, 9-16.
Engel D, Spinello L, Triebel R, Siegwart R, Bülthoff HH and Curio C (May-2009) Medial Features for Superpixel Segmentation, Eleventh IAPR Conference on Machine Vision Applications (MVA 2009), MVA Organizing Committee, Tokyo, Japan, 248-252.
Engel D and Curio C (December-2008) Scale-invariant medial features based on gradient vector flow fields, 19th International Conference on Pattern Recognition (ICPR 2008), IEEE Service Center, Piscataway, NJ, USA, 1-4.
Engel D, Curio C, Tcheang L, Mohler B and Bülthoff HH (October-2008) A psychophysically calibrated controller for navigating through large environments in a limited free-walking space, 15th ACM Symposium on Virtual Reality Software and Technology (VRST 2008), ACM Press, New York, NY, USA, 157-164.

Posters (7):

Neth CT, Souman JL, Engel D, Kloos U, Bülthoff HH and Mohler BJ (October-2010): Velocity-dependent curvature gain and avatar use for Redirected Walking, 2010 Joint Virtual Reality Conference of EuroVR - EGVE - VEC (JVRC 2010), Stuttgart, Germany.
Curio C and Engel D (May-2010): A Computational Mid-Level Vision Approach For Shape-Specific Saliency Detection, 10th Annual Meeting of the Vision Sciences Society (VSS 2010), Naples, FL, USA, Journal of Vision, 10(7) 1160.
Engel D and Curio C (May-2010): Factors Influencing The Detectability Of Pedestrians In Urban Environments, 10th Annual Meeting of the Vision Sciences Society (VSS 2010), Naples, FL, USA, Journal of Vision, 10(7) 1247.
Engel D (February-22-2010): Towards Robust Scene Analysis: A Versatile Mid-level Feature Framework, Symposium "Neural Encoding of Perception and Action", Tübingen, Germany.
Engel D and Curio C (January-2010): Towards robust scene analysis: A versatile mid-level feature framework, 4th International Conference on Cognitive Systems (CogSys 2010), Zürich, Switzerland.
Engel D, Curio C, Mohler B and Bülthoff HH (April-2008): Exploring a large maze in a limited size free-walking space, Cyberwalk Workshop 2008, Tübingen, Germany.
Engel D and Curio C (July-2007): A Biologically Motivated Approach to Human Body Pose Tracking in Clutter, 10th Tübinger Wahrnehmungskonferenz (TWK 2007), Tübingen, Germany.

Theses (1):

Engel D: Shape-Centered Representations: From Features to Applications, Eberhard-Karls-Universität Tübingen, (2011). PhD thesis

Talks (1):

Engel D, Kottler VA, Malisi CU, Röttig M, Schultheiss SJ, Willing EM, Curio C and Bülthoff HH (August-2010) Abstract Talk: Optimizing minimal sketches of visual object categories, 33rd European Conference on Visual Perception, Lausanne, Switzerland, Perception, 39(ECVP Abstract Supplement) 11.

Export as:
BibTeX, XML, Pubman, Edoc, RTF
Last updated: Tuesday, 18.11.2014