Publications with Fulltext

Permanent URI for this collectionhttps://hdl.handle.net/20.500.14288/6

Browse

Filters

Advanced Search

Filter by

Settings

Department: search.filters.department.Department of Computer EngineeringSponsors: search.filters.sponsors.EU

Search Results

Now showing 1 - 10 of 27
  • Thumbnail Image
    PublicationOpen Access
    Self-supervised monocular scene decomposition and depth estimation
    (IEEE Computer Society, 2021) Department of Computer Engineering; N/A; Güney, Fatma; Safadoust, Sadra; Department of Computer Engineering; Koç Üniversitesi İş Bankası Yapay Zeka Uygulama ve Araştırma Merkezi (KUIS AI)/ Koç University İş Bank Artificial Intelligence Center (KUIS AI); College of Engineering; Graduate School of Sciences and Engineering; 187939; N/A
    Self-supervised monocular depth estimation approaches either ignore independently moving objects in the scene or need a separate segmentation step to identify them. We propose MonoDepthSeg to jointly estimate depth and segment moving objects from monocular video without using any ground-truth labels. We decompose the scene into a fixed number of components where each component corresponds to a region on the image with its own transformation matrix representing its motion. We estimate both the mask and the motion of each component efficiently with a shared encoder. We evaluate our method on three driving datasets and show that our model clearly improves depth estimation while decomposing the scene into separately moving components.
  • Thumbnail Image
    PublicationOpen Access
    Overlapping data transfers with computation on GPU with tiles
    (Institute of Electrical and Electronics Engineers (IEEE), 2017) Zhang, Weiqun; Almgren, Ann; Shalf, John; Department of Computer Engineering; Bastem, Burak; Erten, Didem Unat; Master Student; Faculty Member; Department of Computer Engineering; Graduate School of Sciences and Engineering; N/A; 219274
    GPUs are employed to accelerate scientific applications however they require much more programming effort from the programmers particularly because of the disjoint address spaces between the host and the device. OpenACC and OpenMP 4.0 provide directive based programming solutions to alleviate the programming burden however synchronous data movement can create a performance bottleneck in fully taking advantage of GPUs. We propose a tiling based programming model and its library that simplifies the development of GPU programs and overlaps the data movement with computation. The programming model decomposes the data and computation into tiles and treats them as the main data transfer and execution units, which enables pipelining the transfers to hide the transfer latency. Moreover, partitioning application data into tiles allows the programmer to still take advantage of GPU even though application data cannot fit into the device memory. The library leverages C++ lambda functions, OpenACC directives, CUDA streams and tiling API from TiDA to support both productivity and performance. We show the performance of the library on a data transfer-intensive and a compute-intensive kernels and compare its speedup against OpenACC and CUDA. The results indicate that the library can hide the transfer latency, handle the cases where there is no sufficient device memory, and achieves reasonable performance.
  • Thumbnail Image
    PublicationOpen Access
    Verifying robustness of event-driven asynchronous programs against concurrency
    (Springer, 2017) Bouajjani A., Emmi M., Enea C.; Department of Computer Engineering; Özkan, Burcu Külahcıoğlu; Taşıran, Serdar; Faculty Member; Department of Computer Engineering; College of Engineering
    We define a correctness criterion, called robustness against concurrency, for a class of event-driven asynchronous programs that are at the basis of modern UI frameworks in Android, iOS, and Javascript. A program is robust when all possible behaviors admitted by the program under arbitrary procedure and event interleavings are admitted even if asynchronous procedures (respectively, events) are assumed to execute serially, one after the other, accessing shared memory in isolation. We characterize robustness as a conjunction of two correctness criteria: event-serializability (i.e., events can be seen as atomic) and event determinism (executions within each event are insensitive to the interleavings between concurrent tasks dynamically spawned by the event). Then, we provide efficient algorithms for checking these two criteria based on polynomial reductions to reachability problems in sequential programs. This result is surprising because it allows to avoid explicit handling of all concurrent executions in the analysis, which leads to an important gain in complexity. We demonstrate via case studies on Android apps that the typical mistakes programmers make are captured as robustness violations, and that violations can be detected efficiently using our approach.
  • Thumbnail Image
    PublicationOpen Access
    PROTEST-ER: retraining BERT for protest event extraction
    (Association for Computational Linguistics (ACL), 2021) Caselli, Tommaso; Basile, Angelo; Department of Sociology; Department of Computer Engineering; Hürriyetoğlu, Ali; Mutlu, Osman; Teaching Faculty; Researcher; Department of Sociology; Department of Computer Engineering; College of Social Sciences and Humanities; College of Engineering
    We analyze the effect of further pre-training BERT with different domain specific data as an unsupervised domain adaptation strategy for event extraction. Portability of event extraction models is particularly challenging, with large performance drops affecting data on the same text genres (e.g., news). We present PROTEST-ER, a retrained BERT model for protest event extraction. PROTEST-ER outperforms a corresponding generic BERT on out-of-domain data of 8.1 points. Our best performing models reach 51.91-46.39 F1 across both domains.
  • Thumbnail Image
    PublicationOpen Access
    Verifiable dynamic searchable encryption
    (TÜBİTAK, 2019) Department of Computer Engineering; Etemad, Mohammad; Küpçü, Alptekin; PhD Student; Department of Computer Engineering; Graduate School of Sciences and Engineering; N/A; 168060
    Using regular encryption schemes to protect the privacy of the outsourced data implies that the client should sacrifice functionality for security. Searchable symmetric encryption (SSE) schemes encrypt the data in a way that the client can later search and selectively retrieve the required data. Many SSE schemes have been proposed, starting with static constructions, and then dynamic and adaptively secure constructions but usually in the honest-but-curious model. We propose a verifiable dynamic SSE scheme that is adaptively secure against malicious adversaries. Our scheme supports file modification, which is essential for efficiently working with large files, in addition to the ability to add/delete files. While our main construction is proven secure in the random oracle model (ROM), we also present a solution secure in the standard model with full security proof. Our experiments show that our scheme in the ROM performs a search within a few milliseconds, verifies the result in another few milliseconds, and has a proof overhead of 0.01% only. Our standard model solution, while being asymptotically slower, is still practical, requiring only a small client memory (e.g., ≃488 KB) even for a large file collection (e.g., ≃10 GB), and necessitates small tokens (e.g., ≃156 KB for search and ≃362 KB for file operations).
  • Thumbnail Image
    PublicationOpen Access
    Deep generation of 3D articulated models and animations from 2D stick figures
    (Elsevier, 2022) Akman, Alican; Sahillioğlu, Yusuf; Department of Computer Engineering; Sezgin, Tevfik Metin; Faculty Member; Department of Computer Engineering; College of Engineering; 18632
    Generating 3D models from 2D images or sketches is a widely studied important problem in computer graphics. We describe the first method to generate a 3D human model from a single sketched stick figure. In contrast to the existing human modeling techniques, our method does not require a statistical body shape model. We exploit Variational Autoencoders to develop a novel framework capable of transitioning from a simple 2D stick figure sketch, to a corresponding 3D human model. Our network learns the mapping between the input sketch and the output 3D model. Furthermore, our model learns the embedding space around these models. We demonstrate that our network can generate not only 3D models, but also 3D animations through interpolation and extrapolation in the learned embedding space. In addition to 3D human models, we produce 3D horse models in order to show the generalization ability of our framework. Extensive experiments show that our model learns to generate compatible 3D models and animations with 2D sketches.
  • Thumbnail Image
    PublicationOpen Access
    Multifaceted engagement in social interaction with a machine: the JOKER project
    (Institute of Electrical and Electronics Engineers (IEEE), 2018) Devillers, Laurence; Rosset, Sophie; Duplessis, Guillaume Dubuisson; Bechade, Lucile; El Haddad, Kevin; Dupont, Stephane; Deleglise, Paul; Esteve, Yannick; Lailler, Carole; Gilmartin, Emer; Campbell, Nick; Department of Computer Engineering; Erzin, Engin; Yemez, Yücel; Türker, Bekir Berker; Sezgin, Tevfik Metin; Faculty Member; Faculty Member; Faculty Member; Department of Computer Engineering; College of Engineering; Graduate School of Sciences and Engineering; 34503; 107907; N/A; 18632
    This paper addresses the problem of evaluating engagement of the human participant by combining verbal and nonverbal behaviour along with contextual information. This study will be carried out through four different corpora. Four different systems designed to explore essential and complementary aspects of the JOKER system in terms of paralinguistic/linguistic inputs were used for the data collection. An annotation scheme dedicated to the labeling of verbal and non-verbal behavior have been designed. From our experiment, engagement in HRI should be multifaceted.
  • Thumbnail Image
    PublicationOpen Access
    Team Howard Beale at SemEval-2019 task 4: hyperpartisan news detection with BERT
    (Association for Computational Linguistics (ACL), 2019) Dayanık, Erenay; Department of Computer Engineering; Mutlu, Osman; Can, Ozan Arkan; Researcher; Department of Computer Engineering; Graduate School of Sciences and Engineering
    This paper describes our system for SemEval-2019 Task 4: Hyperpartisan News Detection (Kiesel et al., 2019). We use pretrained BERT (Devlin et al., 2018) architecture and investigate the effect of different fine tuning regimes on the final classification task. We show that additional pretraining on news domain improves the performance on the Hyperpartisan News Detection task. Our system1 ranked 8th out of 42 teams with 78.3% accuracy on the held-out test dataset.
  • Thumbnail Image
    PublicationOpen Access
    Cross-context news corpus for protest event-related knowledge base construction
    (Massachusetts Institute of Technology (MIT) Press, 2021) Department of Sociology; N/A; Department of Computer Engineering; Yörük, Erdem; Hürriyetoğlu, Ali; Gürel, Burak; Duruşan, Fırat; Yoltar, Çağrı; Mutlu, Osman; Yüret, Deniz; Faculty Member; Teaching Faculty; Faculty Member; Researcher; Researcher; Faculty Member; Department of Sociology; Department of Computer Engineering; College of Social Sciences and Humanities; Graduate School of Sciences and Engineering; College of Engineering; 28982; N/A; 219277; N/A; N/A; N/A; 179996
    We describe a gold standard corpus of protest events that comprise various local and international English language sources from various countries. The corpus contains document-, sentence-, and token-level annotations. This corpus facilitates creating machine learning models that automatically classify news articles and extract protest event-related information, constructing knowledge bases that enable comparative social and political science studies. For each news source, the annotation starts with random samples of news articles and continues with samples drawn using active learning. Each batch of samples is annotated by two social and political scientists, adjudicated by an annotation supervisor, and improved by identifying annotation errors semi-automatically. We found that the corpus possesses the variety and quality that are necessary to develop and benchmark text classification and event extraction systems in a cross-context setting, contributing to the generalizability and robustness of automated text processing systems. This corpus and the reported results will establish a common foundation in automated protest event collection studies, which is currently lacking in the literature.
  • Thumbnail Image
    PublicationOpen Access
    Audiovisual synchronization and fusion using canonical correlation analysis
    (Institute of Electrical and Electronics Engineers (IEEE), 2007) Department of Computer Engineering; Department of Electrical and Electronics Engineering; Sargın, Mehmet Emre; Yemez, Yücel; Erzin, Engin; Tekalp, Ahmet Murat; Faculty Member; Faculty Member; Faculty Member; Department of Computer Engineering; Department of Electrical and Electronics Engineering; Graduate School of Sciences and Engineering; College of Engineering; N/A; N/A; 34503; 26207
    It is well-known that early integration (also called data fusion) is effective when the modalities are correlated, and late integration (also called decision or opinion fusion) is optimal when modalities are uncorrelated. In this paper, we propose a new multimodal fusion strategy for open-set speaker identification using a combination of early and late integration following canonical correlation analysis (CCA) of speech and lip texture features. We also propose a method for high precision synchronization of the speech and lip features using CCA prior to the proposed fusion. Experimental results show that i) the proposed fusion strategy yields the best equal error rates (EER), which are used to quantify the performance of the fusion strategy for open-set speaker identification, and ii) precise synchronization prior to fusion improves the EER; hence, the best EER is obtained when the proposed synchronization scheme is employed together with the proposed fusion strategy. We note that the proposed fusion strategy outperforms others because the features used in the late integration are truly uncorrelated, since they are output of the CCA analysis.