Temporal Locality Research Articles

DoTA: Unsupervised Detection of Traffic Anomaly in Driving Videos.

Video anomaly detection (VAD) has been extensively studied for static cameras but is much more challenging in egocentric driving videos where the scenes are extremely dynamic. This paper proposes an unsupervised method for traffic VAD based on future object localization. The idea is to predict future locations of traffic participants over a short horizon, and then monitor the accuracy and consistency of these predictions as evidence of an anomaly. Inconsistent predictions tend to indicate an anomaly has occurred or is about to occur. To evaluate our method, we introduce a new large-scale benchmark dataset called Detection of Traffic Anomaly (DoTA)containing 4,677 videos with temporal, spatial, and categorical annotations. We also propose a new VAD evaluation metric, called spatial-temporal area under curve (STAUC), and show that it captures how well a model detects both temporal and spatial locations of anomalies unlike existing metrics that focus only on temporal localization. Experimental results show our method outperforms state-of-the-art methods on DoTA in terms of both metrics. We offer rich categorical annotations in DoTA to benchmark video action detection and online action detection methods. The DoTA dataset has been made available at: https://github.com/MoonBlvd/Detection-of-Traffic-Anomaly.

Sequence as A Whole: A Unified Framework for Video Action Localization with Long-range Text Query.

Comprehensive understanding of video content requires both spatial and temporal localization. However, there lacks a unified video action localization framework, which hinders the coordinated development of this field. Existing 3D CNN methods take fixed and limited input length at the cost of ignoring temporally long-range cross-modal interaction. On the other hand, despite having large temporal context, existing sequential methods often avoid dense cross-modal interactions for complexity reasons. To address this issue, in this paper, we propose a unified framework which handles the whole video in sequential manner with long-range and dense visual-linguistic interaction in an end-to-end manner. Specifically, a lightweight relevance filtering based transformer (Ref-Transformer) is designed, which is composed of relevance filtering based attention and temporally expanded MLP. The text-relevant spatial regions and temporal clips in video can be efficiently highlighted through the relevance filtering and then propagated among the whole video sequence with the temporally expanded MLP. Extensive experiments on three sub-tasks of referring video action localization, i.e., referring video segmentation, temporal sentence grounding, and spatiotemporal video grounding, show that the proposed framework achieves the state-of-the-art performance in all referring video action localization tasks.

M2D2: Maximum-Mean-Discrepancy Decoder for Temporal Localization of Epileptic Brain Activities.

Recent years have seen growing interest in leveraging deep learning models for monitoring epilepsy patients based on electroencephalographic (EEG) signals. However, these approaches often exhibit poor generalization when applied outside of the setting in which training data was collected. Furthermore, manual labeling of EEG signals is a time-consuming process requiring expert analysis, making fine-tuning patient-specific models to new settings a costly proposition. In this work, we propose the Maximum-Mean-Discrepancy Decoder (M2D2) for automatic temporal localization and labeling of seizures in long EEG recordings to assist medical experts. We show that M2D2 achieves 76.0% and 70.4% of F1-score for temporal localization when evaluated on EEG data gathered in a different clinical setting than the training data. The results demonstrate that M2D2 yields substantially higher generalization performance than other state-of-the-art deep learning-based approaches.

Accelerating Parallel Applications Based on Graph Reordering for Random Network Topologies

The Message Passing Interface (MPI) is a crucial programming tool for enabling communication between processes in parallel applications. The goal of MPI users is to allocate tasks to processors in a way that maximizes both spatial and temporal locality in the network. However, this can be challenging, especially in large-scale networks where maximizing processor locality may not be feasible at runtime. To address this issue, we propose the use of <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Hamorder</i> , an offline node reassignment approach that takes into account physical processor locations based on graph reordering for <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Random</i> network topologies. <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Hamorder</i> aims to optimize task mapping for improved performance in parallel applications, whether for multiple tasks or within a single task. Additionally, we investigate the potential of improving MPI applications through runtime parameter tuning based on <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Hamorder</i> . Our evaluation results show that <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Hamorder</i> provides a 27.3% improvement in performance compared to the <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Gorder</i> algorithm on <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Random</i> topologies, which is a state-of-the-art solution designed with the aim of enhancing cache locality and achieves this goal by rearranging the vertices of a graph in a way that places the vertices that are typically accessed together in close proximity. Moreover, our autotuning framework using <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Hamorder</i> results in an average speedup of 1.38x for targeted MPI applications by searching through various runtime parameter combinations.

Distributed Self-Adjusting Tree Networks

The performance of many data-centric cloud applications critically depends on the performance of the underlying datacenter network. Reconfigurable optical technologies have recently introduced a novel opportunity to improve datacenter network performance, by allowing to dynamically adjust the network topology according to the demand. However, the vision of self-adjusting networks raises the fundamental question how such networks can be efficiently operated in a scalable and distributed manner. This article presents <inline-formula><tex-math notation="LaTeX">$DiSplayNet$</tex-math></inline-formula> , the first fully distributed self-adjusting network. <inline-formula><tex-math notation="LaTeX">$DiSplayNet$</tex-math></inline-formula> relies on algorithms that perform decentralized and concurrent topological adjustments to account for changes in the demand. We propose two natural metrics to evaluate the performance of distributed self-adjusting networks, the <i>amortized work</i> (the cost of routing on and adjusting the network) and the <i>makespan</i> (the time it takes to serve a set of communication requests). We present a rigorous formal analysis of the work and makespan of <inline-formula><tex-math notation="LaTeX">$DiSplayNet$</tex-math></inline-formula> , which can be seen as an interesting generalization of analyses known from sequential self-adjusting datastructures. We complement our theoretical contribution with an extensive trace-driven simulation study, shedding light on the opportunities and limitations of leveraging spatial and temporal locality and concurrency in self-adjusting networks.

Концепт «педагогическое событие» в языковой картине образовательной действительности

In order to clarify the terminological apparatus of modern pedagogy, the article discusses the concept of “pedagogical event” and its relationship with the main pedagogical categories: education, training and upbringing. On the basis of research conducted within the framework of linguistics, philosophy, psychology, integral signs of the “event” were identified, namely, the significance of changes in the life of the subject, spatial and temporal localization, which allows to define the “pedagogical event” in a dynamic aspect. If the purpose of education is understood as the transmission, reproduction and generation of culture, then a pedagogical event as a significant change for a person can be defined as a change through the adaptation of the cultural experience of humanity into the cultural experience of a student, as the translation of culture from one person to another.

Temporal Uncertainty Localization to Enable Human-in-the-Loop Analysis of Dynamic Contrast-Enhanced Cardiac MRI Datasets.

Dynamic contrast-enhanced (DCE) cardiac magnetic resonance imaging (CMRI) is a widely used modality for diagnosing myocardial blood flow (perfusion) abnormalities. During a typical free-breathing DCE-CMRI scan, close to 300 time-resolved images of myocardial perfusion are acquired at various contrast "wash in/out" phases. Manual segmentation of myocardial contours in each time-frame of a DCE image series can be tedious and time-consuming, particularly when non-rigid motion correction has failed or is unavailable. While deep neural networks (DNNs) have shown promise for analyzing DCE-CMRI datasets, a "dynamic quality control" (dQC) technique for reliably detecting failed segmentations is lacking. Here we propose a new space-time uncertainty metric as a dQC tool for DNN-based segmentation of free-breathing DCE-CMRI datasets by validating the proposed metric on an external dataset and establishing a human-in-the-loop framework to improve the segmentation results. In the proposed approach, we referred the top 10% most uncertain segmentations as detected by our dQC tool to the human expert for refinement. This approach resulted in a significant increase in the Dice score (p < 0.001) and a notable decrease in the number of images with failed segmentation (16.2% to 11.3%) whereas the alternative approach of randomly selecting the same number of segmentations for human referral did not achieve any significant improvement. Our results suggest that the proposed dQC framework has the potential to accurately identify poor-quality segmentations and may enable efficient DNN-based analysis of DCE-CMRI in a human-in-the-loop pipeline for clinical interpretation and reporting of dynamic CMRI datasets.

Expression of RSPH6A in the first wave of rat spermatogenesis and oxidative stress conditions: Attenuation by melatonin.

Here, we report, for the first time, the temporal expression and localization of axonemal radial spoke head homolog A (RSPH6A) protein during the first wave of rat spermatogenesis and in oxidative stress conditions. For the developmental study, testes were collected from rats at different developmental stages (7, 14, 21, 28, 35, 42, and 60 postnatal days); for in vivo treatment, 24 rats were treated with cadmium and/or melatonin. From each sample, western blot (WB) and immunofluorescence (IF) analyses for RSPH6A were performed. RSPH6A expression starts at 21 PND alongside the appearance of I spermatocytes (SPC) with a significant increase up to 60 PND. Data were confirmed by IF analysis, showing that RPSH6A expression is restricted to I and II SPC, spermatids, and mature sperm. In vivo experiments showed that the expression and localization of RSPH6A in the testis and epididymal spermatozoa of adult rats treated with cadmium were impaired. Interestingly, melatonin (an antioxidant), given together with Cd, can counteract its damaging effects. All combined data confirm that RSPH6A contributes to the onset of fertility by acting on sperm motility, raising the possibility of using RSPH6A as a marker for normal fertility in the general population.

Multi-Dimensional Attention With Similarity Constraint for Weakly-Supervised Temporal Action Localization

Multi-Dimensional Attention With Similarity Constraint for Weakly-Supervised Temporal Action Localization

ІСТОРИЧНИЙ І МОВНО-КУЛЬТУРНИЙ ВИМІРИ НОБЕЛЬСЬКОГО ЄВАНГЕЛІЯ 1520 РОКУ

A manuscript from the beginning of the 16th century from Noble (Pinsk County, which is now in Rivne oblast, Ukraine) was studied, representing the religious writing of the Grand Duchy of Lithuania in historical and language-cultural aspects. The afterword to the Nobel Gospel was analyzed, and the historical figures named in it were identified (Prince Fedor Ivanovich Yaroslavich, Bishop Jonah). Information about them was collected based on historical sources and scientific literature. The ‘Ruthenian’ edition of the Church Slavonic language, manifestations of the second South Slavic influence in the text, is characterized. The monument’s history information, now represented in the Library of the Vrublevsky Academy of Sciences of Lithuania (Lithuania), was systematized. The Nobel Gospel, which has not been the subject of scientific study until now, was analyzed during an interdisciplinary study. The historical context of the monument’s creation based on the afterword is analyzed. Its temporal and spatial localization, the possible sphere of operation, ancient storage of the rarity, scribes (Chivs), and historians were established. Prince Fedor Ivanovich Yaroslavich carried out charitable activities. He founded and supported churches and monasteries in the Pinsk District. The bishops of Pinsk and Turov, Vasian, and later Jonah, mentioned in the afterword, appealed to the king and received a “salary” letter stating that it should be their will and approval for the construction of churches and monasteries. The analyzed manuscript from Polissia was proved to reveal the book traditions of the 16th century and the eccentricity of the centers of rewriting texts of religious writing. The main manifestations of the second South Slavic influence in the orthography of the manuscript are characterized. Different orthographic and phonetic features of the ‘Ruthenian’ edition of the Church Slavonic language have been analyzed. Since the 60s of the 19th century, the manuscript from Polissia was transported to the Vilnius Public Library. The Manuscript Department was founded here, and the first descriptions of the monument were compiled.

Step by Step: A Gradual Approach for Dense Video Captioning

Dense video captioning aims to localize and describe events for storytelling in untrimmed videos. It is a conceptually very challenging task that requires concise, relevant, and coherent captioning based on high-quality event localization. Unlike simple temporal action localization tasks without overlapping events, dense video captioning requires detecting multiple/overlapping regions in order to branch out the video story. Most existing methods generate numerous candidate event proposals and then eliminate duplicate ones using a event proposal selection algorithm (e.g., non-maximum suppression) or generate event proposals directly through box prediction and binary classification mechanisms, similar to object detection tasks. Despite these efforts, the aforementioned approaches tend to fail to localize overlapping events into different stories, hindering high-quality captioning. In this paper, we propose SBS, a dense video captioning framework with a gradual approach that addresses the challenge of localizing overlapping events and eventually constructs high-quality captioning. SBS accurately estimates the number of explicit events for each video snippet and then detects the boundaries context/activities, which are the details for generating the event proposals. Based on both the number of events and boundaries, SBS generates the event proposals. SBS encodes the context of the event sequence and finally generates sentences describing the event proposals. Our framework is fairly effective in localizing multiple/overlapping events, thus experimental results show the state-of-the-art performance compared to the existing methods.

Meaning formations of the personality of high school students living in various conditions of the new Russian reality

Introduction. Meaning formations are the main constitutive characteristics of a personality in early adolescence; they determine the dominant tendencies of its development, activity, attitude towards the world, other people and oneself. Today, the issue of meaning formations in young people in various socio-economic and cultural-historical conditions is becoming urgent. The research aims to analyze certain types and properties of meaning formations (meaningful orientations in life, personal meanings, meaning of life) of Altai and Lugansk schoolchildren. Research methods. The study was carried out within the project of the Ministry of Science and Higher Education, the Ministry of Education and the Russian Movement of Children and Youth ‘University Shifts’. The sample consisted of 320 people studying in Grades 10-11 (secondary school) in Altai Krai and Lugansk People's Republic. Life meaning orientations, types and properties of life meanings, and a life meaning crisis were diagnosed. The following methods were used: ‘Test of Life Meaning Orientations’ by D. A. Leontyev, ‘System of Life Meanings’ by V. Yu. Kotlyakov, ‘Questionnaire of Life Meaning Crisis’ by K. V. Karpinsky, and questionnaire ‘On the Meaning of Life’ by V. E. Chudnovsky. Methods of mathematical and statistical analysis were Mann-Whitney U test for two independent samples, Spearman's correlation, and multiple regression analysis. Research results. The lack of meaningfulness of life among Altai (19%) and Lugansk (49%) high school students is associated with weak temporal localization of meanings in relation to the future and present, and with insufficient internal localization of control (21% and 2%). The past primarily acts as a source of meaning generation. The lack of harmony in meaning formations is expressed in a meaning crisis in high school students of both research groups of a moderate (33% and 36%) and high degree (44% and 45%). The meaning formations of the personality of Altai high school students are characterized by the dominance of existential (38%) and hedonistic meanings (35%) with a relatively even distribution of types of meanings according to their orientation. For Lugansk schoolchildren, the leading ones were status meanings (39%), communicative meanings (34%) and self-realization meanings (32%). According to the orientation criterion, egocentric meanings dominate (57%). Conclusion. Specific properties of semantic formations (localization, harmony, and hierarchy) among Altai and Lugansk high school students were revealed, and differences were found in the hierarchical structure of life meanings and in predictors of meaningfulness of life and life meaning crisis. The results of the study can help organize psychological and pedagogical support for high school students experiencing life changes.

Perfect in the Old Uighur Language

The article discusses the semantic nature of the Turkic perfect, its semantic zone limitations and possible grammar tools of expressing Perfect in the Old Uighur language. Goals. The study aims to determine grammar tools of expressing Perfect meanings in the Old Uighur language. Morphological tools that are able to perform the meanings included into the semantic zone of Perfect are being studied. Perfect and description of its semantic zone are connected with the concepts of aspect, actionality and ways of verbal action. In this regard, the issue of the relevance of these terms, in relation to the Old Uighur perfect, is under discussion. The grammar tools expressing Perfect are studied according to the point of view that excludes Aspect as a grammar category in the Turkic languages. Studying aspectual and temporal meanings of Perfect diachronically on the basis of the Od Uighur material may give a hint to understand the Turkic perfect and its expressions in genetically related modern languages. Functional-semantic and typological methods are used to identify and describe morphological tools with a perfect meaning. Scripts of the Old Uighur language, dated back to 10th–11th centuries, are of biographical and religious-philosophical content. Work with scripts, comprehension of existing theoretical approaches to understanding Perfect as an element of aspectual functional-semantic field show that Perfect meanings in the Old Uighur language are expressed by morphological tools that can be correlated with models. The study concludes that means expressing the Perfect meaning in the Old Uighur language are able to mark temporal localization and/or aspect.

A computational modeling framework for pre-clinical evaluation of cardiac mapping systems.

There are a variety of difficulties in evaluating clinical cardiac mapping systems, most notably the inability to record the transmembrane potential throughout the entire heart during patient procedures which prevents the comparison to a relevant "gold standard". Cardiac mapping systems are comprised of hardware and software elements including sophisticated mathematical algorithms, both of which continue to undergo rapid innovation. The purpose of this study is to develop a computational modeling framework to evaluate the performance of cardiac mapping systems. The framework enables rigorous evaluation of a mapping system's ability to localize and characterize (i.e., focal or reentrant) arrhythmogenic sources in the heart. The main component of our tool is a library of computer simulations of various dynamic patterns throughout the entire heart in which the type and location of the arrhythmogenic sources are known. Our framework allows for performance evaluation for various electrode configurations, heart geometries, arrhythmias, and electrogram noise levels and involves blind comparison of mapping systems against a "silver standard" comprised of computer simulations in which the precise transmembrane potential patterns throughout the heart are known. A feasibility study was performed using simulations of patterns in the human left atria and three hypothetical virtual catheter electrode arrays. Activation times (AcT) and patterns (AcP) were computed for three virtual electrode arrays: two basket arrays with good and poor contact and one high-resolution grid with uniform spacing. The average root mean squared difference of AcTs of electrograms and those of the nearest endocardial action potential was less than 1ms and therefore appears to be a poor performance metric. In an effort to standardize performance evaluation of mapping systems a novel performance metric is introduced based on the number of AcPs identified correctly and those considered spurious as well as misclassifications of arrhythmia type; spatial and temporal localization accuracy of correctly identified patterns was also quantified. This approach provides a rigorous quantitative analysis of cardiac mapping system performance. Proof of concept of this computational evaluation framework suggests that it could help safeguard that mapping systems perform as expected as well as provide estimates of system accuracy.

Optimistic Data Parallelism for FPGA-Accelerated Sketching

Sketches are a popular approximation technique for large datasets and high-velocity data streams. While custom FPGA-based hardware has shown admirable throughput at sketching, the state-of-the-art exploits data parallelism by fully replicating resources and constructing independent summaries for every parallel input value. We consider this approach pessimistic, as it guarantees constant processing rates by provisioning resources for the worst case. We propose a novel optimistic sketching architecture for FPGAs that partitions a single sketch into multiple independent banks shared among all input values, thus significantly reducing resource consumption. However, skewed input data distributions can result in conflicting accesses to banks and impair the processing rate. To mitigate the effect of skew, we add mergers that exploit temporal locality by combining recent updates. Our evaluation shows that an optimistic architecture is feasible and reduces the utilization of critical FPGA resources proportionally to the number of parallel input values. We further show that FPGA accelerators provide up to 2.6 x higher throughput than a recent CPU and GPU, while larger sketch sizes enabled by optimistic architectures improve accuracy by up to an order of magnitude in a realistic sketching application.

Utilizing Temporal Locality for Hash Tables with Circular Chaining

A hash table with separate chaining typically adopts linearly linked lists as buckets to resolve collisions. This study demonstrates that converting bucket structures from linear to circular chaining enables buckets to utilize temporal locality and improve search performance. Unlike linear chaining, circular chaining can track the most recently accessed entry and preserve the reachability of all bucket entries without complicated data structures and operations. We defined temporal locality interval (TLI) to represent the period during which subsequent bucket access repeats itself on a single entry. We analyzed the average search cost using the TLI length and load factor. The average search cost converges to the minimum when the TLI length dominates the load factor. In our experiments using the SPEC CPU 2006 benchmark suite, circular chaining manifested 1.14 comparisons, reducing the cost of linear chaining by 45.71% when the load factor was 0.99. The improvement is notable, particularly for tables with a high load factor and uneven distribution of bucket sizes.

Towards better utilization of pseudo labels for weakly supervised temporal action localization

Weakly supervised temporal action localization (WS-TAL) aims to simultaneously recognize and localize action instances of interest in untrimmed videos with the use of the video-level label only. Some works have demonstrated that pseudo labels play an important role for performance improvement in WS-TAL. Since pseudo labels are inevitably inaccurate, direct adoption of noisy labels can lead to inappropriate knowledge transfer. Although some previous studies have shown the benefits of using only “reliable” pseudo labels, performance improvement is still limited. In this work, we experimentally analyze how the noise in pseudo labels affects model performance within the self-distillation framework. Motivated by the finding that incorrect pseudo labels with large confidence scores have a significant impact on performance, we propose the overconfidence suppression (OCS) strategy to mitigate the effect of the overconfident pseudo labels, and thus prevent over-fitting of the student model. In addition, a simplified contrast learning method is utilized to fine-tune the feature representation by increasing the separation of the foreground and background snippets. Equipped with the proposed methods, the benefits of pseudo labels can be better exploited and allow the model to achieve state-of-the-art performance on THUMOS’14 and ActivityNet-1.2 benchmarks.

Temporal action detection with dynamic weights based on curriculum learning

To enable temporal action localization, the computer needs to recognize the locations and classes of action instances in a video. The main challenge to temporal action detection is that the videos are often long and untrimmed, consisting of varying action content. Existing temporal action detection frameworks exhibit a gap between the training and testing phases, which is detrimental to model performance. Specifically, all positive samples are trained identically in the training phase. By contrast, in the testing phase, the positive samples with the best classification and localization scores are selected, while all others are suppressed. To mitigate this issue, we build an auxiliary branch to unify the training and testing procedures. In the construction of the auxiliary branch, we design a dynamic weighting strategy based on curriculum learning, where the weights of training samples are a combination of their classification and localization scores. Motivated by the speculation of curriculum learning, we emphasize the importance of classification and localization scores in different training stages. The classification score accounts for a higher proportion of the combined score in the early stages of the training process. As the epoch increases, the localization score gradually increases in proportion as well. The experimental results demonstrate that our methodology of curriculum-based learning enhances the performance of current action localization techniques. On THUMOS14, our technique outperforms the existing state-of-the-art technique (57.6% vs 55.5%). And the performance on ActivityNet v1.3 (mAP@Avg) reaches 35.4%.

Amplifying action-context greater: image segmentation-guided intraoperative active bleeding detection

ABSTRACT The intraoperative active bleeding (iAB) detection model can be used for image-guided surgery and as a significant statistical index in predicting patient outcomes after surgery. However, detecting iAB is difficult due to the similarity between active and non-active bleeding or active bleeding in a small area. Using the spatial and temporal characteristics of the iAB area within frames simultaneously can overcome this. We propose a novel training method that can adequately fuse image segmentation and temporal action localisation models for effective iAB detection. The proposed active bleeding detection model has the following supervision process: First, annotate temporal localisation information for active bleeding that is relatively easy to annotate. Next, in the active bleeding section, where temporal localisation is annotated, spatial localisation information for selected frames is annotated and used as auxiliary information for active bleeding detection. We constructed a cross-validation set of 40 robotic subtotal gastrectomies and verified the ability of an active bleeding model guided by image segmentation information to bring improvements to the active bleeding recognition task. In addition, we applied performance evaluation for outcome analysis by measuring errors in iAB duration and counting in surgical videos for each algorithm. 1

MaPHeA: A Framework for Lightweight Memory Hierarchy-aware Profile-guided Heap Allocation

Hardware performance monitoring units (PMUs) are a standard feature in modern microprocessors, providing a rich set of microarchitectural event samplers. Recently, numerous profile-guided optimization (PGO) frameworks have exploited them to feature much lower profiling overhead compared to conventional instrumentation-based frameworks. However, existing PGO frameworks mainly focus on optimizing the layout of binaries; they overlook rich information provided by the PMU about data access behaviors over the memory hierarchy. Thus, we propose MaPHeA, a lightweight M emory hierarchy- a ware P rofile-guided He ap A llocation framework applicable to both HPC and embedded systems. MaPHeA guides and applies the optimized allocation of dynamically allocated heap objects with very low profiling overhead and without additional user intervention to improve application performance. To demonstrate the effectiveness of MaPHeA, we apply it to optimizing heap object allocation in an emerging DRAM-NVM heterogeneous memory system (HMS), selective huge-page utilization, and controlling the cacheability of the objects with the low temporal locality. In an HMS, by identifying and placing frequently accessed heap objects to the fast DRAM region, MaPHeA improves the performance of memory-intensive graph-processing and Redis workloads by 56.0% on average over the default configuration that uses DRAM as a hardware-managed cache of slow NVM. By identifying large heap objects that cause frequent TLB misses and allocating them to huge pages, MaPHeA increases the performance of the read and update operations of Redis by 10.6% over the transparent huge-page implementation of Linux. Also, by distinguishing the objects that cause cache pollution due to their low temporal locality and applying write-combining to them, MaPHeA improves the performance of STREAM and RADIX workloads by 20.0% on average over the system without cacheability control.