Czech Technical University in Prague

As a present to Mizar on its 50th anniversary, we develop an AI/TP system that automatically proves about 60% of the Mizar theorems in the hammer setting. We also automatically prove 75% of the Mizar theorems when the automated provers are helped by using only the premises used in the human-written Mizar proofs. We describe the methods and large-scale experiments leading to these results. This includes in particular the E and Vampire provers, their ENIGMA and Deepire learning modifications, a number of learning-based premise selection methods, and the incremental loop that interleaves growing a corpus of millions of ATP proofs with training increasingly strong AI/TP systems on them. We also present a selection of Mizar problems that were proved automatically.

Author(s): Collaboration, The ATLAS; Aad, G; Abat, E; Abdallah, J; Abdelalim, AA; Abdesselam, A; Abdinov, O; Abi, BA; Abolins, M; Abramowicz, H; Acerbi, E; Acharya, BS; Achenbach, R; Ackers, M; Adams, DL; Adamyan, F; Addy, TN; Aderholz, M; Adorisio, C; Adragna, P; Aharrouche, M; Ahlen, SP; Ahles, F; Ahmad, A; Ahmed, H; Aielli, G; Åkesson, PF; Åkesson, TPA; Akimov, AV; Alam, SM; Albert, J; Albrand, S; Aleksa, M; Aleksandrov, IN; Aleppo, M; Alessandria, F; Alexa, C; Alexander, G; Alexopoulos, T; Alimonti, G; Aliyev, M; Allport, PP; Allwood-Spiers, SE; Aloisio, A; Alonso, J; Alves, R; Alviggi, MG; Amako, K; Amaral, P; Amaral, SP; Ambrosini, G; Ambrosio, G; Amelung, C; Ammosov, VV; Amorim, A; Amram, N; Anastopoulos, C; Anderson, B; Anderson, KJ; Anderssen, EC; Andreazza, A; Andrei, V; Andricek, L; Andrieux, M-L; Anduaga, XS; Anghinolfi, F; Antonaki, A; Antonelli, M; Antonelli, S; Apsimon, R; Arabidze, G; Aracena, I; Arai, Y; Arce, ATH; Archambault, JP; Arguin, J-F; Arik, E; Arik, M; Arms, KE; Armstrong, SR; Arnaud, M; Arnault, C; Artamonov, A; Asai, S; Ask, S

This paper investigates long-term tracking of unknown objects in a video stream. The object is defined by its location and extent in a single frame. In every frame that follows, the task is to determine the object's location and extent or indicate that the object is not present. We propose a novel tracking framework (TLD) that explicitly decomposes the long-term tracking task into tracking, learning, and detection. The tracker follows the object from frame to frame. The detector localizes all appearances that have been observed so far and corrects the tracker if necessary. The learning estimates the detector's errors and updates it to avoid these errors in the future. We study how to identify the detector's errors and learn from them. We develop a novel learning method (P-N learning) which estimates the errors by a pair of "experts": (1) P-expert estimates missed detections, and (2) N-expert estimates false alarms. The learning process is modeled as a discrete dynamical system and the conditions under which the learning guarantees improvement are found. We describe our real-time implementation of the TLD framework and the P-N learning. We carry out an extensive quantitative evaluation which shows a significant improvement over state-of-the-art approaches.

Technological evolution of mobile user equipment (TIEs), such as smartphones or laptops, goes hand-in-hand with evolution of new mobile applications. However, running computationally demanding applications at the TIEs is constrained by limited battery capacity and energy consumption of the TIEs. A suitable solution extending the battery life-time of the TIEs is to offload the applications demanding huge processing to a conventional centralized cloud. Nevertheless, this option introduces significant execution delay consisting of delivery of the off loaded applications to the cloud and back plus time of the computation at the cloud. Such a delay is inconvenient and makes the offloading unsuitable for real-time applications. To cope with the delay problem, a new emerging concept, known as mobile edge computing (MEC), has been introduced. The MEC brings computation and storage resources to the edge of mobile network enabling it to run the highly demanding applications at the TIE while meeting strict delay requirements. The MEC computing resources can be exploited also by operators and third parties for specific purposes. In this paper, we first describe major use cases and reference scenarios where the MEC is applicable. After that we survey existing concepts integrating MEC functionalities to the mobile networks and discuss current advancement in standardization of the MEC. The core of this survey is, then, focused on user-oriented use case in the MEC, i.e., computation offloading. In this regard, we divide the research on computation offloading to three key areas: 1) decision on computation offloading; 2) allocation of computing resource within the MEC; and 3) mobility management. Finally, we highlight lessons learned in area of the MEC and we discuss open research challenges yet to be addressed in order to fully enjoy potentials offered by the MEC.

Self-driving vehicles are a maturing technology with the potential to reshape mobility by enhancing the safety, accessibility, efficiency, and convenience of automotive transportation. Safety-critical tasks that must be executed by a self-driving vehicle include planning of motions through a dynamic environment shared with other vehicles and pedestrians, and their robust executions via feedback control. The objective of this paper is to survey the current state of the art on planning and control algorithms with particular regard to the urban setting. A selection of proposed techniques is reviewed along with a discussion of their effectiveness. The surveyed approaches differ in the vehicle mobility model used, in assumptions on the structure of the environment, and in computational requirements. The side by side comparison presented in this survey helps to gain insight into the strengths and limitations of the reviewed approaches and assists with system level design choices.

The Large Hadron Collider (LHC) at CERN will extend the frontiers of particle physics with its
\nunprecedented high energy and luminosity. Inside the LHC, bunches of up to 1011 protons (p)
\nwill collide 40 million times per second to provide 14 TeV proton-proton collisions at a design
\nluminosity of 1034 cm􀀀2s􀀀1. The LHC will also collide heavy ions (A), in particular lead nuclei, at
\n5.5 TeV per nucleon pair, at a design luminosity of 1027 cm􀀀2s􀀀1.
\nThe high interaction rates, radiation doses, particle multiplicities and energies, as well as the
\nrequirements for precision measurements have set new standards for the design of particle detectors.
\nTwo general purpose detectors, ATLAS (A Toroidal LHC ApparatuS) and CMS (Compact
\nMuon Solenoid) have been built for probing p-p and A-A collisions.
\nThis paper presents a comprehensive overview of the ATLAS detector prior to the first LHC
\ncollisions, written as the installation of the ATLAS detector is nearing completion. This detector
\nrepresents the work of a large collaboration of several thousand physicists, engineers, technicians,
\nand students over a period of fifteen years of dedicated design, development, fabrication, and installation.

ALICE (A Large Ion Collider Experiment) is a general-purpose, heavy-ion detector at the CERN LHC which focuses on QCD, the strong-interaction sector of the Standard Model. It is designed to address the physics of strongly interacting matter and the quark-gluon plasma at extreme values of energy density and temperature in nucleus-nucleus collisions. Besides running with Pb ions, the physics programme includes collisions with lighter ions, lower energy running and dedicated proton-nucleus runs. ALICE will also take data with proton beams at the top LHC energy to collect reference data for the heavy-ion programme and to address several QCD topics for which ALICE is complementary to the other LHC detectors. The ALICE detector has been built by a collaboration including currently over 1000 physicists and engineers from 105 Institutes in 30 countries, Its overall dimensions are 16 x 16 x 26 m(3) with a total weight of approximately 10 000 t. The experiment consists of 18 different detector systems each with its own specific technology choice and design constraints, driven both by the physics requirements and the experimental conditions expected at LHC. The most stringent design constraint is to cope with the extreme particle multiplicity anticipated in central Pb-Pb collisions. The different subsystems were optimized to provide high-momentum resolution as well as excellent Particle Identification (PID) over a broad range in momentum, up to the highest multiplicities predicted for LHC. This will allow for comprehensive studies of hadrons, electrons, muons, and photons produced in the collision of heavy nuclei. Most detector systems are scheduled to be installed and ready for data taking by mid-2008 when the LHC is scheduled to start operation, with the exception of parts of the Photon Spectrometer (PHOS), Transition Radiation Detector (TRD) and Electro Magnetic Calorimeter (EMCal). These detectors will be completed for the high-luminosity ion run expected in 2010. This paper describes in detail the detector components as installed for the first data taking in the summer of 2008.

We tackle the problem of large scale visual place recognition, where the task is to quickly and accurately recognize the location of a given query photograph. We present the following three principal contributions. First, we develop a convolutional neural network (CNN) architecture that is trainable in an end-to-end manner directly for the place recognition task. The main component of this architecture, NetVLAD, is a new generalized VLAD layer, inspired by the "Vector of Locally Aggregated Descriptors" image representation commonly used in image retrieval. The layer is readily pluggable into any CNN architecture and amenable to training via backpropagation. Second, we develop a training procedure, based on a new weakly supervised ranking loss, to learn parameters of the architecture in an end-to-end manner from images depicting the same places over time downloaded from Google Street View Time Machine. Finally, we show that the proposed architecture significantly outperforms non-learnt image representations and off-the-shelf CNN descriptors on two challenging place recognition benchmarks, and improves over current state-of-the-art compact image representations on standard image retrieval benchmarks.

Results of the ICDAR 2015 Robust Reading Competition are presented. A new Challenge 4 on Incidental Scene Text has been added to the Challenges on Born-Digital Images, Focused Scene Images and Video Text. Challenge 4 is run on a newly acquired dataset of 1,670 images evaluating Text Localisation, Word Recognition and End-to-End pipelines. In addition, the dataset for Challenge 3 on Video Text has been substantially updated with more video sequences and more accurate ground truth data. Finally, tasks assessing End-to-End system performance have been introduced to all Challenges. The competition took place in the first quarter of 2015, and received a total of 44 submissions. Only the tasks newly introduced in 2015 are reported on. The datasets, the ground truth specification and the evaluation protocols are presented together with the results and a brief summary of the participating methods.

The state of the art in visual object retrieval from large databases is achieved by systems that are inspired by text retrieval. A key component of these approaches is that local regions of images are characterized using high-dimensional descriptors which are then mapped to ldquovisual wordsrdquo selected from a discrete vocabulary.This paper explores techniques to map each visual region to a weighted set of words, allowing the inclusion of features which were lost in the quantization stage of previous systems. The set of visual words is obtained by selecting words based on proximity in descriptor space. We describe how this representation may be incorporated into a standard tf-idf architecture, and how spatial verification is modified in the case of this soft-assignment. We evaluate our method on the standard Oxford Buildings dataset, and introduce a new dataset for evaluation. Our results exceed the current state of the art retrieval performance on these datasets, particularly on queries with poor initial recall where techniques like query expansion suffer. Overall we show that soft-assignment is always beneficial for retrieval with large vocabularies, at a cost of increased storage requirements for the index.

Abstract The wide-baseline stereo problem, i.e. the problem of establishing correspondences between a pair of images taken from different viewpoints is studied. A new set of image elements that are put into correspondence, the so called extremal regions , is introduced. Extremal regions possess highly desirable properties: the set is closed under (1) continuous (and thus projective) transformation of image coordinates and (2) monotonic transformation of image intensities. An efficient (near linear complexity) and practically fast detection algorithm (near frame rate) is presented for an affinely invariant stable subset of extremal regions, the maximally stable extremal regions (MSER). A new robust similarity measure for establishing tentative correspondences is proposed. The robustness ensures that invariants from multiple measurement regions (regions obtained by invariant constructions from extremal regions), some that are significantly larger (and hence discriminative) than the MSERs, may be used to establish tentative correspondences. The high utility of MSERs, multiple measurement regions and the robust metric is demonstrated in wide-baseline experiments on image pairs from both indoor and outdoor scenes. Significant change of scale (3.5×), illumination conditions, out-of-plane rotation, occlusion, locally anisotropic scale change and 3D translation of the viewpoint are all present in the test problems. Good estimates of epipolar geometry (average distance from corresponding points to the epipolar line below 0.09 of the inter-pixel distance) are obtained.

A measurement of the Higgs boson mass is presented based on the combined data samples of the ATLAS and CMS experiments at the CERN LHC in the H→γγ and H→ZZ→4ℓ decay channels. The results are obtained from a simultaneous fit to the reconstructed invariant mass peaks in the two channels and for the two experiments. The measured masses from the individual channels and the two experiments are found to be consistent among themselves. The combined measured mass of the Higgs boson is m_{H}=125.09±0.21 (stat)±0.11 (syst) GeV.

Short-term tracking is an open and challenging problem for which discriminative correlation filters (DCF) have shown excellent performance.We introduce the channel and spatial reliability concepts to DCF tracking and provide a novel learning algorithm for its efficient and seamless integration in the filter update and the tracking process. The spatial reliability map adjusts the filter support to the part of the object suitable for tracking. This both allows to enlarge the search region and improves tracking of non-rectangular objects. Reliability scores reflect channel-wise quality of the learned filters and are used as feature weighting coefficients in localization. Experimentally, with only two simple standard features, HoGs and Colornames, the novel CSRDCF method - DCF with Channel and Spatial Reliability - achieves state-of-the-art results on VOT 2016, VOT 2015 and OTB100. The CSR-DCF runs in real-time on a CPU.

In this work we address the problem of finding reliable pixel-level correspondences under difficult imaging conditions. We propose an approach where a single convolutional neural network plays a dual role: It is simultaneously a dense feature descriptor and a feature detector. By postponing the detection to a later stage, the obtained keypoints are more stable than their traditional counterparts based on early detection of low-level structures. We show that this model can be trained using pixel correspondences extracted from readily available large-scale SfM reconstructions, without any further annotations. The proposed method obtains state-of-the-art performance on both the difficult Aachen Day-Night localization dataset and the InLoc indoor localization benchmark, as well as competitive performance on other benchmarks for image matching and 3D reconstruction.

A new robust matching method is proposed. The progressive sample consensus (PROSAC) algorithm exploits the linear ordering defined on the set of correspondences by a similarity function used in establishing tentative correspondences. Unlike RANSAC, which treats all correspondences equally and draws random samples uniformly from the full set, PROSAC samples are drawn from progressively larger sets of top-ranked correspondences. Under the mild assumption that the similarity measure predicts correctness of a match better than random guessing, we show that PROSAC achieves large computational savings. Experiments demonstrate it is often significantly faster (up to more than hundred times) than RANSAC. For the derived size of the sampled set of correspondences as a function of the number of samples already drawn, PROSAC converges towards RANSAC in the worst case. The power of the method is demonstrated on wide-baseline matching problems.

Combined ATLAS and CMS measurements of the Higgs boson production and decay rates, as well as constraints on its couplings to vector bosons and fermions, are presented. The combination is based on the analysis of five production processes, namely gluon fusion, vector boson fusion, and associated production with a W or a Z boson or a pair of top quarks, and of the six decay modes H → ZZ, W W , γγ, ττ, bb, and μμ. All results are reported assuming a value of 125.09 GeV for the Higgs boson mass, the result of the combined measurement by the ATLAS and CMS experiments. The analysis uses the CERN LHC proton-proton collision data recorded by the ATLAS and CMS experiments in 2011 and 2012, corresponding to integrated luminosities per experiment of approximately 5 fb$^{−1}$ at $\sqrt{s}$=7 TeV and 20 fb−1 at $\sqrt{s}$=8 TeV. The Higgs boson production and decay rates measured by the two experiments are combined within the context of three generic parameterisations: two based on cross sections and branching fractions, and one on ratios of coupling modifiers. Several interpretations of the measurements with more model-dependent parameterisations are also given. The combined signal yield relative to the Standard Model prediction is measured to be 1.09 ± 0.11. The combined measurements lead to observed significances for the vector boson fusion production process and for the H → ττ decay of 5.4 and 5.5 standard deviations, respectively. The data are consistent with the Standard Model predictions for all parameterisations considered.

This paper shows that the performance of a binary classifier can be significantly improved by the processing of structured unlabeled data, i.e. data are structured if knowing the label of one example restricts the labeling of the others. We propose a novel paradigm for training a binary classifier from labeled and unlabeled examples that we call P-N learning. The learning process is guided by positive (P) and negative (N) constraints which restrict the labeling of the unlabeled set. P-N learning evaluates the classifier on the unlabeled data, identifies examples that have been classified in contradiction with structural constraints and augments the training set with the corrected samples in an iterative process. We propose a theory that formulates the conditions under which P-N learning guarantees improvement of the initial classifier and validate it on synthetic and real data. P-N learning is applied to the problem of on-line learning of object detector during tracking. We show that an accurate object detector can be learned from a single example and an unlabeled video sequence where the object may occur. The algorithm is compared with related approaches and state-of-the-art is achieved on a variety of objects (faces, pedestrians, cars, motorbikes and animals).

Learning text-video embeddings usually requires a dataset of video clips with\nmanually provided captions. However, such datasets are expensive and time\nconsuming to create and therefore difficult to obtain on a large scale. In this\nwork, we propose instead to learn such embeddings from video data with readily\navailable natural language annotations in the form of automatically transcribed\nnarrations. The contributions of this work are three-fold. First, we introduce\nHowTo100M: a large-scale dataset of 136 million video clips sourced from 1.22M\nnarrated instructional web videos depicting humans performing and describing\nover 23k different visual tasks. Our data collection procedure is fast,\nscalable and does not require any additional manual annotation. Second, we\ndemonstrate that a text-video embedding trained on this data leads to\nstate-of-the-art results for text-to-video retrieval and action localization on\ninstructional video datasets such as YouCook2 or CrossTask. Finally, we show\nthat this embedding transfers well to other domains: fine-tuning on generic\nYoutube videos (MSR-VTT dataset) and movies (LSMDC dataset) outperforms models\ntrained on these datasets alone. Our dataset, code and models will be publicly\navailable at: www.di.ens.fr/willow/research/howto100m/.\n

This paper presents a method for detection of steganographic methods that embed in the spatial domain by adding a low-amplitude independent stego signal, an example of which is least significant bit (LSB) matching. First, arguments are provided for modeling the differences between adjacent pixels using first-order and second-order Markov chains. Subsets of sample transition probability matrices are then used as features for a steganalyzer implemented by support vector machines. The major part of experiments, performed on four diverse image databases, focuses on evaluation of detection of LSB matching. The comparison to prior art reveals that the presented feature set offers superior accuracy in detecting LSB matching. Even though the feature set was developed specifically for spatial domain steganalysis, by constructing steganalyzers for ten algorithms for JPEG images, it is demonstrated that the features detect steganography in the transform domain as well.

We report the first measurement of charged particle elliptic flow in Pb-Pb collisions at sqrt[S(NN)] =2.76 TeV with the ALICE detector at the CERN Large Hadron Collider. The measurement is performed in the central pseudorapidity region (|η|<0.8) and transverse momentum range 0.2<p t<5.0 GeV/c. The elliptic flow signal v₂, measured using the 4-particle correlation method, averaged over transverse momentum and pseudorapidity is 0.087 ± 0.002(stat) ± 0.003(syst) in the 40%-50% centrality class. The differential elliptic flow v₂ p t reaches a maximum of 0.2 near p t =3 GeV/c. Compared to RHIC Au-Au collisions at sqrt[S(NN)] 200 GeV, the elliptic flow increases by about 30%. Some hydrodynamic model predictions which include viscous corrections are in agreement with the observed increase.