Red Carpet to Fight Club: Partially-supervised Domain Transfer for Face Recognition in Violent Videos

2021-01-01
Bilge, Yunus Can
Yucel, Mehmet Kerim
Cinbiş, Ramazan Gökberk
İKİZLER CİNBİŞ, NAZLI
DUYGULU ŞAHİN, PINAR
In many real-world problems, there is typically a large discrepancy between the characteristics of data used in training versus deployment. A prime example is the analysis of aggression videos: in a criminal incidence, typically suspects need to be identified based on their clean portrait-like photos, instead of their prior video recordings. This results in three major challenges; large domain discrepancy between violence videos and ID-photos, the lack of video examples for most individuals and limited training data availability. To mimic such scenarios, we formulate a realistic domain-transfer problem, where the goal is to transfer the recognition model trained on clean posed images to the target domain of violent videos, where training videos are available only for a subset of subjects. To this end, we introduce the "WildestFaces" dataset, tailored to study cross-domain recognition under a variety of adverse conditions. We divide the task of transferring a recognition model from the domain of clean images to the violent videos into two sub-problems and tackle them using (i) stacked affine-transforms for classifier-transfer, (ii) attention-driven pooling for temporal-adaptation. We additionally formulate a self-attention based model for domain-transfer. We establish a rigorous evaluation protocol for this "clean-to-violent" recognition task, and present a detailed analysis of the proposed dataset and the methods. Our experiments highlight the unique challenges introduced by the WildestFaces dataset and the advantages of the proposed approach.
IEEE Winter Conference on Applications of Computer Vision (WACV)

Suggestions

SEADATANET - A PAN-EUROPEAN INFRASTRUCTURE FOR OCEAN AND MARINE DATA MANAGEMENT
Bingel, Ferit(2011-3-31)
Data availability is of vital importance for marine research but most of the European data are fragmented, not always validated and not easily accessible. In the 40 countries bordering the European seas, more than 600 scientific laboratories from governmental organizations and private industry collect data by using various sensors on board of research vessels, submarines, fixed and drifting platforms, airplanes and satellites to measure physical, geophysical, geological, biological and chemical parameters, ...
Domain adaptation on graphs by learning graph topologies: theoretical analysis and an algorithm
Vural, Elif (The Scientific and Technological Research Council of Turkey, 2019-01-01)
Traditional machine learning algorithms assume that the training and test data have the same distribution, while this assumption does not necessarily hold in real applications. Domain adaptation methods take into account the deviations in data distribution. In this work, we study the problem of domain adaptation on graphs. We consider a source graph and a target graph constructed with samples drawn from data manifolds. We study the problem of estimating the unknown class labels on the target graph using the...
CLASS IMBALANCE AND ACTIVE LEARNING
Attenberg, Josh; Ertekin Bolelli, Şeyda (2013-01-01)
The performance of a predictive model is tightly coupled with the data used during training. While using more examples in the training will often result in a better informed, more accurate model; limits on computer memory and real-world costs associated with gathering labeled examples often constrain the amount of data that can be used for training. In settings where the number of training examples is limited, it often becomes meaningful to carefully see just which examples are selected. In active learning ...
Design and implementation of index structures for fuzzy spatial databases
Sozer, Aziz; Yazıcı, Adnan (2007-07-01)
Over the years database management systems have evolved to include spatially referenced data. Because spatial data are complex and have a number of unique constraints (i.e., spatial components and uncertain properties), spatial database systems can be effective only if the spatial data are properly handled at the physical level. Therefore, it is important to develop an effective spatial and aspatial indexing technique to facilitate flexible spatial and/or aspatial querying for such databases. For this purpo...
Index structures for flexible querying in fuzzy spatial databases
Sozer, A; Yazıcı, Adnan (2004-07-29)
Database systems can be effective only if the data are properly handled at the physical level. Therefore, it is important to develop an effective spatial and aspatial indexing technique to facilitate flexible spatial and/or aspatial querying for spatial databases. In this study we adapt a number of spatial index structures, such as Multi-level grid file (MLGF), G-tree, R-tree, and R*-tree, for fuzzy spatial databases and compare the performances of these structures for various flexible queries.
Citation Formats
Y. C. Bilge, M. K. Yucel, R. G. Cinbiş, N. İKİZLER CİNBİŞ, and P. DUYGULU ŞAHİN, “Red Carpet to Fight Club: Partially-supervised Domain Transfer for Face Recognition in Violent Videos,” presented at the IEEE Winter Conference on Applications of Computer Vision (WACV), ELECTR NETWORK, 2021, Accessed: 00, 2021. [Online]. Available: https://hdl.handle.net/11511/92597.