2017 IEEE International Conference on Software Testing, Verification and Validation (ICST)
Android has become one of the most popular operating systems for mobile devices. As the number of... more Android has become one of the most popular operating systems for mobile devices. As the number of applications for the Android ecosystem grows, so is their complexity, increasing the need for runtime verification on the Android platform. Unfortunately, despite the presence of several runtime verification frameworks for Java bytecode, DEX bytecode used in Android does not benefit from such a wide support. While a few runtime verification tools support applications developed for Android, such tools offer only limited bytecode coverage and may not be able to detect property violations in certain classes. In this paper, we show that ADRENALIN-RV, our new runtime verification tool for Android, overcomes this limitation. In contrast to other frameworks, ADRENALIN-RV weaves monitoring code at load time and is able to instrument all loaded classes. In addition to the default classes inside the application package (APK), ADRENALIN-RV covers both the Android class library and libraries dynamically loaded from the storage, network, or generated dynamically, which related tools cannot verify. Evaluation results demonstrate the increased code coverage of ADRENALIN-RV with respect to other runtime validation tools for Android. Thanks to ADRENALIN-RV, we were able to detect violations that cannot be detected by other tools.
Selon cette invention, un support lisible par ordinateur contient un logiciel qui, lorsqu'il ... more Selon cette invention, un support lisible par ordinateur contient un logiciel qui, lorsqu'il est lu par un ordinateur, pousse l'ordinateur a mettre en oeuvre un procede de surveillance base sur un site etendu. Le procede consiste a recevoir des donnees de surveillance, y compris des vues cibles, d'une pluralite de capteurs situes au niveau d'un site; a synchroniser les donnees de surveillance avec une seule source temporelle; a maintenir un modele de site du site, lequel modele de site comprend une carte de site, une carte a l'echelle humaine et un modele de reseau de capteurs; a analyser les donnees synchronisees a l'aide du modele de site pour determiner si les vues cibles representent un meme objet physique dans le site; a creer une carte cible correspondant a un objet physique dans le site, laquelle carte cible comprend au moins une vue cible; a recevoir un evenement d'interet global defini par l'utilisateur, lequel evenement d'interet global ...
Runtime verification (RV) is a field of study which suffers from a lack of dedicated benchmarks. ... more Runtime verification (RV) is a field of study which suffers from a lack of dedicated benchmarks. Many published evaluations of RV tools rely on workloads which are not representative of real-world programs. In this paper, we present a methodology to automatically discover relevant open-source projects for evaluating RV tools. This is done by analyzing unit tests on a large number of projects hosted on GitHub. Our evaluation shows that analyzing a large number of open-source projects—instead of a handful of manually selected workloads—provides better insight into the behavior of three state-of-the-art RV tools (JavaMOP, MarQ, and Muffin) based on two metrics (memory utilization and runtime overhead). By monitoring test executions of a large number of projects, we show that none of the evaluated RV tools wins for both metrics.
Selon l'invention, une camera video peut surveiller une zone controlee a partir d'une pos... more Selon l'invention, une camera video peut surveiller une zone controlee a partir d'une position possible quelconque. Un module d'estimation du flux d'objets controle la direction de deplacement des objets situes dans la zone controlee, et peut separer les objets se deplacant de maniere coherente des autres objets. Un module d'estimation du nombre d'objets peut calculer la densite des objets ( par ex., une foule de personnes). Un module de classification des densites d'objets peut classer les densites pour les affecter a des categories personnalisables.
BACKGROUND Ankle fractures are one of the most common fractures in adults aged 20-65 years. The B... more BACKGROUND Ankle fractures are one of the most common fractures in adults aged 20-65 years. The British Orthopaedic Association (BOA) and British Orthopaedic Foot and Ankle Society (BOFAS) jointly produced Standards for Trauma (BOAST) BOAST 12, with the aim of reducing morbidity by standardising care of these injuries. The primary aim of the AUGMENT study was to determine the extent and clinical effect of variation from BOAST 12. METHODS AUGMENT was a multi-centre prospective trainee led audit of consecutive patients presenting with an ankle fracture within a four-week period. Data were collected on patient demographics, comorbidities, management and 12-week outcome. The BOAST 12 standards were divided into four subgroups; documentation, imaging, management and follow-up. Percentage compliance with each subgroup was analysed. A multivariate logistic regression analysis was used to determine impact of overall compliance on likelihood of discharge in follow-up period. FINDINGS 971 patients were included across 52 sites. The overall rate of BOAST 12 compliance was 41.7%. Variations in practice were observed in clinical documentation, especially of neurovascular status, (40.7%) and VTE assessment (61.5%). Patient management compliance with all 16 of the BOAST 12 standards was associated with a higher rate of discharge during the 12-week follow-up period (p = 0.005). CONCLUSION AUGMENT has demonstrated that the management of ankle fractures is variable across the UK. Over half of patients had aspects of their care that were not BOAST 12 compliant. When compliance was observed, it was associated with earlier discharge from orthopaedic care.
This study compared outcomes of surgical versus conservative management of ankle fractures in adu... more This study compared outcomes of surgical versus conservative management of ankle fractures in adults through a systematic review and meta-analysis. Methods We searched CINAHL, EMBASE, MEDLINE and CENTRAL databases (1946 to June 2019) for randomised and quasi-randomised controlled trials comparing surgical versus conservative management of closed adult ankle fractures of any type. Estimates of effect were pooled using random effects meta-analysis. Results 1153 patients from 7 trials were included. Our primary outcome, ankle function score, was not statistically significantly different at 6-months (pooled mean difference (surgical minus conservative) = 1.0; 95% CI:-2.3 to 4.3; p=0.55) or 12-months or more (pooled mean difference = 4.6; 95% CI:-1.0 to 10.2; p=0.11) between surgical and conservative groups in three trials assessing displaced or unstable fractures, and two trials using non-validated questionnaires. One trial assessing AO-type-B1 fractures without talar shift had a statistically significant difference favouring conservative management, which was not clinically meaningful. Surgery had lower rates of early treatment failure and malunion/non-union, but higher rates of further surgery and infection. Conclusions Surgical and conservative management of displaced or unstable ankle fractures produce similar short-term functional outcomes. The higher risk of early treatment failure and malunion/non-union in the conservative group versus higher rates of further surgery and infection in the surgical group should be considered. Trials are needed to assess longer-term results and inform management of select patient groups.
International Journal of Computers and Applications, 2002
In this paper, we present a method to remove commercials from talk and game show videos and to se... more In this paper, we present a method to remove commercials from talk and game show videos and to segment these videos into host and guest shots. In our approach, we mainly rely on information contained in shot transitions, rather than analyzing the scene content of individual frames. We utilize the inherent difference in scene structure of commercials and talk shows to differentiate between them. Similarly, we make use of the well-defined structure of talk shows, which can be exploited to classify shots as host or guest shots. The entire show is first segmented into camera shots based on color histogram. Then, we construct a data-structure (shot connectivity graph) which links similar shots over time. Analysis of the shot connectivity graph helps us to automatically separate commercials from program segments. This is done by first detecting stories, and then assigning a weight to each story based on its likelihood of being a commercial. Further analysis on stories is done to distinguish shots of the hosts from shots of the guests. We have tested our approach on several fulllength shows (including commercials) and have achieved video segmentation with high accuracy. The whole scheme is fast and works even on low quality video (160x120 pixel images at 5 Hz).
Background: End colostomy rates following colorectal resection vary across institutions in high-i... more Background: End colostomy rates following colorectal resection vary across institutions in high-income settings, being influenced by patient, disease, surgeon and system factors. This study aimed to assess global variation in end colostomy rates after left-sided colorectal resection. Methods: This study comprised an analysis of GlobalSurg-1 and-2 international, prospective, observational cohort studies (2014, 2016), including consecutive adult patients undergoing elective or emergency left-sided colorectal resection within discrete 2-week windows. Countries were grouped into high-, middle-and low-income tertiles according to the United Nations Human Development Index (HDI). Factors associated with colostomy formation versus primary anastomosis were explored using a multilevel, multivariable logistic regression model. Results: In total, 1635 patients from 242 hospitals in 57 countries undergoing left-sided colorectal resection were included: 113 (6⋅9 per cent) from low-HDI, 254 (15⋅5 per cent) from middle-HDI and 1268 (77⋅6 per cent) from high-HDI countries. There was a higher proportion of patients with perforated disease (57⋅5, 40⋅9 and 35⋅4 per cent; P < 0⋅001) and subsequent use of end colostomy (52⋅2, 24⋅8 and 18⋅9 per cent; P < 0⋅001) in low-compared with middle-and high-HDI settings. The association with colostomy use in low-HDI settings persisted (odds ratio (OR) 3⋅20, 95 per cent c.i. 1⋅35 to 7⋅57; P = 0⋅008) after risk adjustment for malignant disease (OR 2⋅34, 1⋅65 to 3⋅32; P < 0⋅001), emergency surgery (OR 4⋅08, 2⋅73 to 6⋅10; P < 0⋅001), time to operation at least 48 h (OR 1⋅99, 1⋅28 to 3⋅09; P = 0⋅002) and disease perforation (OR 4⋅00, 2⋅81 to 5⋅69; P < 0⋅001). Conclusion: Global differences existed in the proportion of patients receiving end stomas after left-sided colorectal resection based on income, which went beyond case mix alone.
Nowadays, many programming language implementations and programming frameworks target the Java Vi... more Nowadays, many programming language implementations and programming frameworks target the Java Virtual Machine (JVM). Examples include the Java and Scala compilers, Oracle’s Truffle framework and the interpreters built on top of it for a variety of dynamic programming languages, as well as big-data frameworks such as Apache Spark, Apache Flink, and Apache Storm. Unfortunately, the JVM provides only limited support for runtime monitoring. The JVM Tool Interface (JVMTI) offers only a very low-level programming model, often introduces high overhead, and does not guarantee proper isolation of the monitoring logic from the observed program. Aspect-Oriented Programming (AOP), in particular AspectJ, is often used to implement runtime monitoring tools. While offering a convenient programming model, the use of such technologies acerbates performance- and isolation-related problems. In this paper, we advocate the use of our dynamic program analysis framework DiSL for runtime monitoring on the JVM. DiSL reconciles an AOP-based programming model, full coverage of all executed bytecodes, optimizations of the monitoring logic, and support for isolation of the monitoring logic. Moreover, DiSL also offers an API to deploy, adapt, and remove monitoring logic at runtime, and it provides seamless support for monitoring also applications running on Android devices.
Many runtime verification tools for the Java virtual machine rely on aspect-oriented programming,... more Many runtime verification tools for the Java virtual machine rely on aspect-oriented programming, particularly on AspectJ, to weave the verification logic into the observed program. However, AspectJ imposes several limitations on the verification tools, such as a restricted join point model and the inability of weaving certain classes, particularly the Java and Android class libraries. In this paper, we show that our domain-specific aspect language DiSL can overcome these limitations. While offering a programming model akin to AspectJ, DiSL features an extensible join point model and ensures weaving with complete bytecode coverage for Java and Android. We present a new compiler that translates runtime-verification aspects written in AspectJ to DiSL. Hence, it is possible to use existing, unmodified runtime verification tools on top of the DiSL framework to bypass the limitations of AspectJ. As a case study, we show that the AspectJ-based runtime verification tool JavaMOP significantly benefits from the automated translation of AspectJ to DiSL code, gaining increased code coverage. Thanks to DiSL, JavaMOP analyses are able to unveil violations in the Java class library that cannot be detected when using AspectJ.
2012 IEEE Conference on Computer Vision and Pattern Recognition, 2012
Low-level appearance as well as spatio-temporal features , appropriately quantized and aggregated... more Low-level appearance as well as spatio-temporal features , appropriately quantized and aggregated into Bagof-Words (BoW) descriptors, have been shown to be effective in many detection and recognition tasks. However, their efficacy for complex event recognition in unconstrained videos have not been systematically evaluated. In this paper, we use the NIST TRECVID Multimedia Event Detection (MED11 [1]) open source dataset, containing annotated data for 15 high-level events, as the standardized test bed for evaluating the low-level features. This dataset contains a large number of user-generated video clips. We consider 7 different low-level features, both static and dynamic, using BoW descriptors within an SVM approach for event detection. We present performance results on the 15 MED11 events for each of the features as well as their combinations using a number of early and late fusion strategies and discuss their strengths and limitations.
2004 International Conference on Image Processing, 2004. ICIP '04.
We propose a solution to the problem of object recognition given a continuous video sequence cont... more We propose a solution to the problem of object recognition given a continuous video sequence containing multiple views of an object. Initially, object models are acquired from images of the objects taken from different views. Recognition is achieved from the video sequences by employing a multiple hypothesis approach. Appearance similarity, and pose transition smoothness constraints are used to estimate the probability of the measurement being generated from a certain model hypothesis at each time instant. A smooth gradient direction feature that is quasi-invariant to illumination changes and noise is used to represent the appearance of object. The pose of the object at each time instant is modelled as a von Mises-Fisher distribution. Recognition is achieved by choosing the hypothesis set that has accumulated the maximum evidence at the end of the sequence. We have performed detailed experiments demonstrating the viability of the proposed approach.
2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'05)
When viewed from a system of multiple cameras with nonoverlapping fields of view, the appearance ... more When viewed from a system of multiple cameras with nonoverlapping fields of view, the appearance of an object in one camera view is usually very different from its appearance in another camera view due to the differences in illumination, pose and camera parameters. In order to handle the change in observed colors of an object as it moves from one camera to another, we show that all brightness transfer functions from a given camera to another camera lie in a low dimensional subspace and demonstrate that this subspace can be used to compute appearance similarity. In the proposed approach, the system learns the subspace of intercamera brightness transfer functions in a training phase during which object correspondences are assumed to be known. Once the training is complete, correspondences are assigned using the maximum a posteriori (MAP) estimation framework using both location and appearance cues. We evaluate the proposed method under several real world scenarios obtaining encouraging results.
2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'05)
Boosting based detection methods have successfully been used for robust detection of faces and pe... more Boosting based detection methods have successfully been used for robust detection of faces and pedestrians. However, a very large amount of labeled examples are required for training such a classifier. Moreover, once trained, the boosted classifier cannot adjust to the particular scenario in which it is employed. In this paper, we propose a co-training based approach to continuously label incoming data and use it for online update of the boosted classifier that was initially trained from a small labeled example set. The main contribution of our approach is that it is an online procedure in which separate views (features) of the data are used for co-training, while the combined view (all features) is used to make classification decisions in a single boosted framework. The features used for classification are derived from Principal Component Analysis of the appearance templates of the training examples. In order to speed up the classification, background modeling is used to prune away stationary regions in an image. Our experiments indicate that starting from a classifier trained on a small training set, significant performance gains can be made through online updation from the unlabeled data.
Proceedings of the British Machine Vision Conference 2014, 2014
We present an algorithm to estimate depth in dynamic video scenes. We propose to learn and infer ... more We present an algorithm to estimate depth in dynamic video scenes. We propose to learn and infer depth in videos from appearance, motion, occlusion boundaries, and geometric context of the scene. Using our method, depth can be estimated from unconstrained videos with no requirement of camera pose estimation, and with significant background/foreground motions. We start by decomposing a video into spatio-temporal regions. For each spatio-temporal region, we learn the relationship of depth to visual appearance, motion, and geometric classes. Then we infer the depth information of new scenes using piecewise planar parametrization estimated within a Markov random field (MRF) framework by combining appearance to depth learned mappings and occlusion boundary guided smoothness constraints. Subsequently, we perform temporal smoothing to obtain temporally consistent depth maps. We present a thorough evaluation of our algorithm on our new dataset and the publicly available Make3d static image dataset.
2008 Second ACM/IEEE International Conference on Distributed Smart Cameras, 2008
Page 1. AUTOMATED VISUAL ANALYSIS IN LARGE SCALE SENSOR NETWORKS Z. Rasheed, X. Cao, K. Shafique,... more Page 1. AUTOMATED VISUAL ANALYSIS IN LARGE SCALE SENSOR NETWORKS Z. Rasheed, X. Cao, K. Shafique, H. Liu, L. Yu, M. Lee, K. Ramnath, T. Choe, O. Javed, and N. Haering Center for Video Understanding Excellence ...
2003 International Conference on Multimedia and Expo. ICME '03. Proceedings (Cat. No.03TH8698), 2003
In this paper, we present a wide area surveiliance system that detects, tracks and classifies mov... more In this paper, we present a wide area surveiliance system that detects, tracks and classifies moving objects across mul-tiple cameras. At the single camera level, tracking is per-formed using a voting based approach that utilizes color and shape cues to establish correspondence. ...
2014 IEEE Conference on Computer Vision and Pattern Recognition, 2014
Detecting pedestrians at a distance from large-format wide-area imagery is a challenging problem ... more Detecting pedestrians at a distance from large-format wide-area imagery is a challenging problem because of low ground sampling distance (GSD) and low frame rate of the imagery. In such a scenario, the approaches based on appearance cues alone mostly fail because pedestrians are only a few pixels in size. Frame-differencing and optical flow based approaches also give poor detection results due to noise, camera jitter and parallax in aerial videos. To overcome these challenges, we propose a novel approach to extract Multi-scale Intrinsic Motion Structure features from pedestrian's motion patterns for pedestrian detection. The MIMS feature encodes the intrinsic motion properties of an object, which are location, velocity and trajectory-shape invariant. The extracted MIMS representation is robust to noisy flow estimates. In this paper, we give a comparative evaluation of the proposed method and demonstrate that MIMS outperforms the state of the art approaches in identifying pedestrians from low resolution airborne videos.
2017 IEEE International Conference on Software Testing, Verification and Validation (ICST)
Android has become one of the most popular operating systems for mobile devices. As the number of... more Android has become one of the most popular operating systems for mobile devices. As the number of applications for the Android ecosystem grows, so is their complexity, increasing the need for runtime verification on the Android platform. Unfortunately, despite the presence of several runtime verification frameworks for Java bytecode, DEX bytecode used in Android does not benefit from such a wide support. While a few runtime verification tools support applications developed for Android, such tools offer only limited bytecode coverage and may not be able to detect property violations in certain classes. In this paper, we show that ADRENALIN-RV, our new runtime verification tool for Android, overcomes this limitation. In contrast to other frameworks, ADRENALIN-RV weaves monitoring code at load time and is able to instrument all loaded classes. In addition to the default classes inside the application package (APK), ADRENALIN-RV covers both the Android class library and libraries dynamically loaded from the storage, network, or generated dynamically, which related tools cannot verify. Evaluation results demonstrate the increased code coverage of ADRENALIN-RV with respect to other runtime validation tools for Android. Thanks to ADRENALIN-RV, we were able to detect violations that cannot be detected by other tools.
Selon cette invention, un support lisible par ordinateur contient un logiciel qui, lorsqu'il ... more Selon cette invention, un support lisible par ordinateur contient un logiciel qui, lorsqu'il est lu par un ordinateur, pousse l'ordinateur a mettre en oeuvre un procede de surveillance base sur un site etendu. Le procede consiste a recevoir des donnees de surveillance, y compris des vues cibles, d'une pluralite de capteurs situes au niveau d'un site; a synchroniser les donnees de surveillance avec une seule source temporelle; a maintenir un modele de site du site, lequel modele de site comprend une carte de site, une carte a l'echelle humaine et un modele de reseau de capteurs; a analyser les donnees synchronisees a l'aide du modele de site pour determiner si les vues cibles representent un meme objet physique dans le site; a creer une carte cible correspondant a un objet physique dans le site, laquelle carte cible comprend au moins une vue cible; a recevoir un evenement d'interet global defini par l'utilisateur, lequel evenement d'interet global ...
Runtime verification (RV) is a field of study which suffers from a lack of dedicated benchmarks. ... more Runtime verification (RV) is a field of study which suffers from a lack of dedicated benchmarks. Many published evaluations of RV tools rely on workloads which are not representative of real-world programs. In this paper, we present a methodology to automatically discover relevant open-source projects for evaluating RV tools. This is done by analyzing unit tests on a large number of projects hosted on GitHub. Our evaluation shows that analyzing a large number of open-source projects—instead of a handful of manually selected workloads—provides better insight into the behavior of three state-of-the-art RV tools (JavaMOP, MarQ, and Muffin) based on two metrics (memory utilization and runtime overhead). By monitoring test executions of a large number of projects, we show that none of the evaluated RV tools wins for both metrics.
Selon l'invention, une camera video peut surveiller une zone controlee a partir d'une pos... more Selon l'invention, une camera video peut surveiller une zone controlee a partir d'une position possible quelconque. Un module d'estimation du flux d'objets controle la direction de deplacement des objets situes dans la zone controlee, et peut separer les objets se deplacant de maniere coherente des autres objets. Un module d'estimation du nombre d'objets peut calculer la densite des objets ( par ex., une foule de personnes). Un module de classification des densites d'objets peut classer les densites pour les affecter a des categories personnalisables.
BACKGROUND Ankle fractures are one of the most common fractures in adults aged 20-65 years. The B... more BACKGROUND Ankle fractures are one of the most common fractures in adults aged 20-65 years. The British Orthopaedic Association (BOA) and British Orthopaedic Foot and Ankle Society (BOFAS) jointly produced Standards for Trauma (BOAST) BOAST 12, with the aim of reducing morbidity by standardising care of these injuries. The primary aim of the AUGMENT study was to determine the extent and clinical effect of variation from BOAST 12. METHODS AUGMENT was a multi-centre prospective trainee led audit of consecutive patients presenting with an ankle fracture within a four-week period. Data were collected on patient demographics, comorbidities, management and 12-week outcome. The BOAST 12 standards were divided into four subgroups; documentation, imaging, management and follow-up. Percentage compliance with each subgroup was analysed. A multivariate logistic regression analysis was used to determine impact of overall compliance on likelihood of discharge in follow-up period. FINDINGS 971 patients were included across 52 sites. The overall rate of BOAST 12 compliance was 41.7%. Variations in practice were observed in clinical documentation, especially of neurovascular status, (40.7%) and VTE assessment (61.5%). Patient management compliance with all 16 of the BOAST 12 standards was associated with a higher rate of discharge during the 12-week follow-up period (p = 0.005). CONCLUSION AUGMENT has demonstrated that the management of ankle fractures is variable across the UK. Over half of patients had aspects of their care that were not BOAST 12 compliant. When compliance was observed, it was associated with earlier discharge from orthopaedic care.
This study compared outcomes of surgical versus conservative management of ankle fractures in adu... more This study compared outcomes of surgical versus conservative management of ankle fractures in adults through a systematic review and meta-analysis. Methods We searched CINAHL, EMBASE, MEDLINE and CENTRAL databases (1946 to June 2019) for randomised and quasi-randomised controlled trials comparing surgical versus conservative management of closed adult ankle fractures of any type. Estimates of effect were pooled using random effects meta-analysis. Results 1153 patients from 7 trials were included. Our primary outcome, ankle function score, was not statistically significantly different at 6-months (pooled mean difference (surgical minus conservative) = 1.0; 95% CI:-2.3 to 4.3; p=0.55) or 12-months or more (pooled mean difference = 4.6; 95% CI:-1.0 to 10.2; p=0.11) between surgical and conservative groups in three trials assessing displaced or unstable fractures, and two trials using non-validated questionnaires. One trial assessing AO-type-B1 fractures without talar shift had a statistically significant difference favouring conservative management, which was not clinically meaningful. Surgery had lower rates of early treatment failure and malunion/non-union, but higher rates of further surgery and infection. Conclusions Surgical and conservative management of displaced or unstable ankle fractures produce similar short-term functional outcomes. The higher risk of early treatment failure and malunion/non-union in the conservative group versus higher rates of further surgery and infection in the surgical group should be considered. Trials are needed to assess longer-term results and inform management of select patient groups.
International Journal of Computers and Applications, 2002
In this paper, we present a method to remove commercials from talk and game show videos and to se... more In this paper, we present a method to remove commercials from talk and game show videos and to segment these videos into host and guest shots. In our approach, we mainly rely on information contained in shot transitions, rather than analyzing the scene content of individual frames. We utilize the inherent difference in scene structure of commercials and talk shows to differentiate between them. Similarly, we make use of the well-defined structure of talk shows, which can be exploited to classify shots as host or guest shots. The entire show is first segmented into camera shots based on color histogram. Then, we construct a data-structure (shot connectivity graph) which links similar shots over time. Analysis of the shot connectivity graph helps us to automatically separate commercials from program segments. This is done by first detecting stories, and then assigning a weight to each story based on its likelihood of being a commercial. Further analysis on stories is done to distinguish shots of the hosts from shots of the guests. We have tested our approach on several fulllength shows (including commercials) and have achieved video segmentation with high accuracy. The whole scheme is fast and works even on low quality video (160x120 pixel images at 5 Hz).
Background: End colostomy rates following colorectal resection vary across institutions in high-i... more Background: End colostomy rates following colorectal resection vary across institutions in high-income settings, being influenced by patient, disease, surgeon and system factors. This study aimed to assess global variation in end colostomy rates after left-sided colorectal resection. Methods: This study comprised an analysis of GlobalSurg-1 and-2 international, prospective, observational cohort studies (2014, 2016), including consecutive adult patients undergoing elective or emergency left-sided colorectal resection within discrete 2-week windows. Countries were grouped into high-, middle-and low-income tertiles according to the United Nations Human Development Index (HDI). Factors associated with colostomy formation versus primary anastomosis were explored using a multilevel, multivariable logistic regression model. Results: In total, 1635 patients from 242 hospitals in 57 countries undergoing left-sided colorectal resection were included: 113 (6⋅9 per cent) from low-HDI, 254 (15⋅5 per cent) from middle-HDI and 1268 (77⋅6 per cent) from high-HDI countries. There was a higher proportion of patients with perforated disease (57⋅5, 40⋅9 and 35⋅4 per cent; P < 0⋅001) and subsequent use of end colostomy (52⋅2, 24⋅8 and 18⋅9 per cent; P < 0⋅001) in low-compared with middle-and high-HDI settings. The association with colostomy use in low-HDI settings persisted (odds ratio (OR) 3⋅20, 95 per cent c.i. 1⋅35 to 7⋅57; P = 0⋅008) after risk adjustment for malignant disease (OR 2⋅34, 1⋅65 to 3⋅32; P < 0⋅001), emergency surgery (OR 4⋅08, 2⋅73 to 6⋅10; P < 0⋅001), time to operation at least 48 h (OR 1⋅99, 1⋅28 to 3⋅09; P = 0⋅002) and disease perforation (OR 4⋅00, 2⋅81 to 5⋅69; P < 0⋅001). Conclusion: Global differences existed in the proportion of patients receiving end stomas after left-sided colorectal resection based on income, which went beyond case mix alone.
Nowadays, many programming language implementations and programming frameworks target the Java Vi... more Nowadays, many programming language implementations and programming frameworks target the Java Virtual Machine (JVM). Examples include the Java and Scala compilers, Oracle’s Truffle framework and the interpreters built on top of it for a variety of dynamic programming languages, as well as big-data frameworks such as Apache Spark, Apache Flink, and Apache Storm. Unfortunately, the JVM provides only limited support for runtime monitoring. The JVM Tool Interface (JVMTI) offers only a very low-level programming model, often introduces high overhead, and does not guarantee proper isolation of the monitoring logic from the observed program. Aspect-Oriented Programming (AOP), in particular AspectJ, is often used to implement runtime monitoring tools. While offering a convenient programming model, the use of such technologies acerbates performance- and isolation-related problems. In this paper, we advocate the use of our dynamic program analysis framework DiSL for runtime monitoring on the JVM. DiSL reconciles an AOP-based programming model, full coverage of all executed bytecodes, optimizations of the monitoring logic, and support for isolation of the monitoring logic. Moreover, DiSL also offers an API to deploy, adapt, and remove monitoring logic at runtime, and it provides seamless support for monitoring also applications running on Android devices.
Many runtime verification tools for the Java virtual machine rely on aspect-oriented programming,... more Many runtime verification tools for the Java virtual machine rely on aspect-oriented programming, particularly on AspectJ, to weave the verification logic into the observed program. However, AspectJ imposes several limitations on the verification tools, such as a restricted join point model and the inability of weaving certain classes, particularly the Java and Android class libraries. In this paper, we show that our domain-specific aspect language DiSL can overcome these limitations. While offering a programming model akin to AspectJ, DiSL features an extensible join point model and ensures weaving with complete bytecode coverage for Java and Android. We present a new compiler that translates runtime-verification aspects written in AspectJ to DiSL. Hence, it is possible to use existing, unmodified runtime verification tools on top of the DiSL framework to bypass the limitations of AspectJ. As a case study, we show that the AspectJ-based runtime verification tool JavaMOP significantly benefits from the automated translation of AspectJ to DiSL code, gaining increased code coverage. Thanks to DiSL, JavaMOP analyses are able to unveil violations in the Java class library that cannot be detected when using AspectJ.
2012 IEEE Conference on Computer Vision and Pattern Recognition, 2012
Low-level appearance as well as spatio-temporal features , appropriately quantized and aggregated... more Low-level appearance as well as spatio-temporal features , appropriately quantized and aggregated into Bagof-Words (BoW) descriptors, have been shown to be effective in many detection and recognition tasks. However, their efficacy for complex event recognition in unconstrained videos have not been systematically evaluated. In this paper, we use the NIST TRECVID Multimedia Event Detection (MED11 [1]) open source dataset, containing annotated data for 15 high-level events, as the standardized test bed for evaluating the low-level features. This dataset contains a large number of user-generated video clips. We consider 7 different low-level features, both static and dynamic, using BoW descriptors within an SVM approach for event detection. We present performance results on the 15 MED11 events for each of the features as well as their combinations using a number of early and late fusion strategies and discuss their strengths and limitations.
2004 International Conference on Image Processing, 2004. ICIP '04.
We propose a solution to the problem of object recognition given a continuous video sequence cont... more We propose a solution to the problem of object recognition given a continuous video sequence containing multiple views of an object. Initially, object models are acquired from images of the objects taken from different views. Recognition is achieved from the video sequences by employing a multiple hypothesis approach. Appearance similarity, and pose transition smoothness constraints are used to estimate the probability of the measurement being generated from a certain model hypothesis at each time instant. A smooth gradient direction feature that is quasi-invariant to illumination changes and noise is used to represent the appearance of object. The pose of the object at each time instant is modelled as a von Mises-Fisher distribution. Recognition is achieved by choosing the hypothesis set that has accumulated the maximum evidence at the end of the sequence. We have performed detailed experiments demonstrating the viability of the proposed approach.
2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'05)
When viewed from a system of multiple cameras with nonoverlapping fields of view, the appearance ... more When viewed from a system of multiple cameras with nonoverlapping fields of view, the appearance of an object in one camera view is usually very different from its appearance in another camera view due to the differences in illumination, pose and camera parameters. In order to handle the change in observed colors of an object as it moves from one camera to another, we show that all brightness transfer functions from a given camera to another camera lie in a low dimensional subspace and demonstrate that this subspace can be used to compute appearance similarity. In the proposed approach, the system learns the subspace of intercamera brightness transfer functions in a training phase during which object correspondences are assumed to be known. Once the training is complete, correspondences are assigned using the maximum a posteriori (MAP) estimation framework using both location and appearance cues. We evaluate the proposed method under several real world scenarios obtaining encouraging results.
2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'05)
Boosting based detection methods have successfully been used for robust detection of faces and pe... more Boosting based detection methods have successfully been used for robust detection of faces and pedestrians. However, a very large amount of labeled examples are required for training such a classifier. Moreover, once trained, the boosted classifier cannot adjust to the particular scenario in which it is employed. In this paper, we propose a co-training based approach to continuously label incoming data and use it for online update of the boosted classifier that was initially trained from a small labeled example set. The main contribution of our approach is that it is an online procedure in which separate views (features) of the data are used for co-training, while the combined view (all features) is used to make classification decisions in a single boosted framework. The features used for classification are derived from Principal Component Analysis of the appearance templates of the training examples. In order to speed up the classification, background modeling is used to prune away stationary regions in an image. Our experiments indicate that starting from a classifier trained on a small training set, significant performance gains can be made through online updation from the unlabeled data.
Proceedings of the British Machine Vision Conference 2014, 2014
We present an algorithm to estimate depth in dynamic video scenes. We propose to learn and infer ... more We present an algorithm to estimate depth in dynamic video scenes. We propose to learn and infer depth in videos from appearance, motion, occlusion boundaries, and geometric context of the scene. Using our method, depth can be estimated from unconstrained videos with no requirement of camera pose estimation, and with significant background/foreground motions. We start by decomposing a video into spatio-temporal regions. For each spatio-temporal region, we learn the relationship of depth to visual appearance, motion, and geometric classes. Then we infer the depth information of new scenes using piecewise planar parametrization estimated within a Markov random field (MRF) framework by combining appearance to depth learned mappings and occlusion boundary guided smoothness constraints. Subsequently, we perform temporal smoothing to obtain temporally consistent depth maps. We present a thorough evaluation of our algorithm on our new dataset and the publicly available Make3d static image dataset.
2008 Second ACM/IEEE International Conference on Distributed Smart Cameras, 2008
Page 1. AUTOMATED VISUAL ANALYSIS IN LARGE SCALE SENSOR NETWORKS Z. Rasheed, X. Cao, K. Shafique,... more Page 1. AUTOMATED VISUAL ANALYSIS IN LARGE SCALE SENSOR NETWORKS Z. Rasheed, X. Cao, K. Shafique, H. Liu, L. Yu, M. Lee, K. Ramnath, T. Choe, O. Javed, and N. Haering Center for Video Understanding Excellence ...
2003 International Conference on Multimedia and Expo. ICME '03. Proceedings (Cat. No.03TH8698), 2003
In this paper, we present a wide area surveiliance system that detects, tracks and classifies mov... more In this paper, we present a wide area surveiliance system that detects, tracks and classifies moving objects across mul-tiple cameras. At the single camera level, tracking is per-formed using a voting based approach that utilizes color and shape cues to establish correspondence. ...
2014 IEEE Conference on Computer Vision and Pattern Recognition, 2014
Detecting pedestrians at a distance from large-format wide-area imagery is a challenging problem ... more Detecting pedestrians at a distance from large-format wide-area imagery is a challenging problem because of low ground sampling distance (GSD) and low frame rate of the imagery. In such a scenario, the approaches based on appearance cues alone mostly fail because pedestrians are only a few pixels in size. Frame-differencing and optical flow based approaches also give poor detection results due to noise, camera jitter and parallax in aerial videos. To overcome these challenges, we propose a novel approach to extract Multi-scale Intrinsic Motion Structure features from pedestrian's motion patterns for pedestrian detection. The MIMS feature encodes the intrinsic motion properties of an object, which are location, velocity and trajectory-shape invariant. The extracted MIMS representation is robust to noisy flow estimates. In this paper, we give a comparative evaluation of the proposed method and demonstrate that MIMS outperforms the state of the art approaches in identifying pedestrians from low resolution airborne videos.
Uploads
Papers by Omar Javed