Search

Semantic segmentation of point cloud data, an efficient 3-D scattered point representation, is a fundamental task for various applications, such as autonomous driving and 3-D telepresence. In recent years, deep learning techniques have achieved significant progress in semantic segmentation, especially in the 2-D image setting. However, due to the irregularity of point clouds, most of them cannot be applied to this special data representation directly. While recent works are able to handle the irregularity problem and maintain the permutation invariance, most of them fail to capture the valuable high-dimensional local feature in fine granularity. Inspired by fuzzy mathematical methods and the analysis on the drawbacks of current state-of-the-art works, in this article, we propose a novel deep neural model, Fuzzy3DSeg, that is able to directly feed in the point clouds while maintaining invariant to the permutation of the data feeding order. We deeply integrate the learning of the fuzzy neighborhood feature of each point into our network architecture, so as to perform operations on high-dimensional features. We demonstrate the effectiveness of this network architecture level integration, compared with methods of the fuzzy data preprocessing cascading neural network. Comprehensive experiments on two challenging datasets demonstrate that the proposed Fuzzy3DSeg significantly outperforms the state-of-the-art methods.

Many existing solutions perform document-level sentiment classification based on local document only, ignoring other texts that might contribute to better classification accuracy. In this paper, we propose a novel speculative sentiment classification model named SSC. In SSC, we speculate that users with similar rating behaviours are more likely to write documents of similar sentiments toward a product. The motivation of SSC, therefore, is to exploit those speculative similar documents for improving classification accuracy. The proposed SSC model consists of three main components, namely, user-product interaction (UPI) component, document encoding (DE) component, and speculative similar document (SSD) component. The UPI component models user-product interactions, and encodes user/product ratings behaviours into user/product embeddings. The DE component utilizes learned user/product embeddings to capture the informative word vectors for comprising more accurate document representations. The SSD component aggregates documents written by similar users toward the same product for speculative sentiment classification. Because the user similarities are calculated based on user embeddings that encode user rating behaviours, the aggregated documents are more likely to have similar sentiments. The three components are seamlessly integrated into a unified model. In the unified manner, these three components are jointly optimized, and they mutually complement each other to enhance sentiment classification. We conduct extensive experiments on three public datasets, and demonstrate the advantage of the proposed SSC model over state-of-the-art baselines.

Existing research exploits the semantic information from reviews to complement user-item interactions for item recommendation. However, as these approaches either defer the user-item interactions until the prediction layer or simply concatenate all the reviews of a user/item into a single review, they fail to capture the complex correlations between each user-item pair or introduce noises. Thus, we propose a novel Hierarchical and Interactive Gate Network (HIGnet) model for rating prediction. Modeling local word informativeness and global review semantics in a hierarchical manner enable us to exploit textual features of users/items and capture complex semantic user-item correlations at different levels of granularities. Experiments on five challenging real-world datasets demonstrate the state-of-the-art performance of the proposed HIGnet model. To facilitate community research, the implementation of the proposed model is made publicly available (https://github.com/uqjwen/higan).

Unlike traditional video recommendations, micro-video inherits the characteristics of social platforms, such as social relation. A large amount of micro-videos showing explosive growth is badly affecting the user’s choice. In this paper, we propose a multi-source multi-net micro-video recommendation model that recommends micro-videos fitting users’ best interests. Different from existing works, as micro-video inherits the characteristics of social platforms, we simultaneously incorporate multi-source content data of items and multi-networks of users to learn user and item representations for recommendation. This information can be complementary to each other in a way that multi-modality data can bridge the semantic gap among items, while multi-type user networks, such as following and reposting, are able to propagate the preferences among users. Furthermore, to discover the hidden categories of micro-videos that properly match users’ interests, we interactively learn the user–item representations and perform the hidden item category clustering. The resulted categorical representations are interacted with user representations to model user preferences at different levels of hierarchies. Finally, multi-source content item data, multi-type user networks and hidden item categories are jointly modelled in a unified recommender, and the parameters of the model are collaboratively learned to boost the recommendation performance. Experiments on a real dataset demonstrate the effectiveness of the proposed model and its advantage over the state-of-the-art baselines.

In this paper, we propose solutions to advance answer selection in Community Question Answering (CQA). Automatically selecting correct answers can significantly improve intelligence for CQA, as users are not required to browse the large quantity of texts and select the right answers manually. Also, automatic answers selection can minimize the time for satisfying users seeking the correct answers and maximize user engagement with the site. Unlike previous works, we propose a hybrid attention mechanism to model question-answer pairs. Specifically, for each word, we calculate the intra-sentence attention indicating its local importance and the inter-sentence attention implying its importance to the counterpart sentence. The inter-sentence attention is based on the interactions between question-answer pairs, and the combination of these two attention mechanisms enables us to align the most informative parts in question-answer pairs for sentence matching. Additionally, we exploit user information for answer selection due to the fact that users are more likely to provide correct answers in their areas of expertise. We model users from their written answers to alleviate data sparsity problem, and then learn user representations according to the informative parts in sentences that are useful for question-answer matching task. This mean of modelling users can bridge the semantic gap between different users, as similar users may have the same way of wording their answers. The representations of users, questions and answers are learnt in an end-to-end neural network in a mean that best explains the interrelation between question-answer pairs. We validate the proposed model on a public dataset, and demonstrate its advantages over the baselines with thorough experiments.

Tetraphenylethylene-naphthopyran (TPE-NP) and tetraphenylethylene-spirooxazine (TPE-SO) with conjugated photochromic groups and fluorophores were designed and synthesized. Both of them exhibit the comprehensive effects of aggregation-induced emission (AIE) and photochromism while different photochromic groups lead to a unique photochromic rate and fluorescence intensity. Among them, TPE-NP has significant photochromic and fluorescence effects with good stability. Therefore, it could be used as an anti-counterfeiting ink and is expected to be used in fluorescence switches in the future.

Smart hydrogels are regarded as novel sensors due to large reversible volume change in response to external stimuli. Here we prepared a novel visual multi-triggered sensor based on inverse opal hydrogel (IOH) that responses to pH, temperature and light, respectively. The hydrogel was composed of hydrophilic pH and temperature responsive poly(dimethylaminoethyl methacrylate) (PDMAEMA), and light responsive spiropyran-containing segment. When the pH decreased, the sensor changed its structural color from blue to orange-red, and the structural color was reversibly recovered by increasing pH. In addition, the color of sensor can be controlled to change from green to violet while applying a temperature field. Furthermore, the reflection peak of IOH exhibited obvious shifted behavior by UV/visible irradiation. Repeated condition changes revealed that the sensor has a long lifetime.

Community question answering (CQA) services have emerged as a type of popular social platforms. In the social network, experts provide knowledgeable answers to the questions in their domain of expertise, while celebrities publish influential opinions toward the topics led by some questions. Given the large amount of knowledge organized in the format of question–answers, an interesting research problem is to recommend questions to users so as to maximize their engagements with the platform. However, recommending questions in CQA services is a non-trivial task. Data sources in the CQA services are of different types. It is challenging to incorporate heterogeneous information for the recommendation task. Furthermore, data sparsity is an inherent problem in such platforms. In this paper, we propose a model that is able to jointly model both implicit and explicit information for question recommendation. The model integrates multiple data sources and addresses the problem of data heterogeneity. In the proposed model, we dynamically discover latent user groups and incorporate those hierarchical information to bridge the semantic gaps among users in the shared latent space. We evaluate the proposed model on two real-world datasets, and demonstrate that our model outperforms the state-of-the-art alternatives by a large margin. We also investigate different structures of the proposed model to study the effects of different data sources. MOE (Min. of Education, S’pore) Published version

A general and efficient way for enlarging the absorption wavelength of optical materials with high molar absorptivity was explored by incorporating a hydroxyl group ortho-position to a bridging group between the electron-withdrawing and the electron-donating groups. Both theoretical deduction and experimental data indicated that the molecular design strategy was rational and efficacious. Comparing to its analogy, the target materials, 3-hydroxyl-4-(N- ethyl-N-ethoxyl phenylazo) naphthalene sulfonic acid (HPNSA), possessed the improved optical property by extending the intrinsic D-π-A conjugated structure and selective recognition to Fe2+ by forming stable 5-member rings. Importantly, the proposed material was effectively applied for colorimetric determination of Fe2+ in four real environmental water samples with an accurate deviation of 4.0% and a low detection limit of 38 nmol L−1. The action mechanism between HPNSA and Fe2+ was investigated in detail.

Common use of smartphones is a compelling reason for performing activity recognition with on-board sensors as it is more practical than other approaches, such as wearable sensors and augmented environments. Many solutions have been proposed by academia, but practical use is limited to experimental settings. Ad hoc solutions exist with different degrees in recognition accuracy and efficiency. To ease the development of activity recognition for the mobile application eco-system, Google released an activity recognition service on their Android platform. In this paper, we present a systematic evaluation of this activity recognition service and share the lesson learnt. Through our experiments, we identified scenarios in which the recognition accuracy was barely acceptable. We analyze the cause of the inaccuracy and propose four practical and light-weight solutions to significantly improve the recognition accuracy and efficiency. Our evaluation confirmed the improvement. As a contribution, we released the proposed solutions as open-source projects for developers who want to incorporate activity recognition into their applications.

We demonstrate that different people perform activities differently.Combine AdaBoost with LDA to build general activity model with minimum labelled data.Hybrid AdaBoost with HMM&CRF for temporal regulatization of human activities.Use publicly available datasets to validate the proposed methods. Activity recognition has been a hot topic for decades, from the scientific research to the development of off-the-shelf commercial products. Since people perform the activities differently, to avoid overfitting, building a general model with activity data of various users is required before the deployment for personal use. However, annotating a large amount of activity data is expensive and time-consuming. In this paper, we build a general model for activity recognition with a limited amount of labelled data. We combine Latent Dirichlet Allocation (LDA) and AdaBoost to jointly train a general activity model with partially labelled data. After that, when AdaBoost is used for online prediction, we combine it with graphical models (such as HMM and CRF) to exploit the temporal information in human activities to smooth out the accidental misclassifications. Experiments with publicly available datasets show that we are able to obtain the accuracy of more than 90% with 1% labelled data.

In recommender systems, the user-item interaction data is usually sparse and not sufficient for learning comprehensive user/item representations for recommendation. To address this problem, we propose a novel dual-bridging recommendation model (DBRec). DBRec performs latent user/item group discovery simultaneously with collaborative filtering, and interacts group information with users/items for bridging similar users/items. Therefore, a user's preference over an unobserved item, in DBRec, can be bridged by the users within the same group who have rated the item, or the user-rated items that share the same group with the unobserved item. In addition, we propose to jointly learn user-user group (item-item group) hierarchies, so that we can effectively discover latent groups and learn compact user/item representations. We jointly integrate collaborative filtering, latent group discovering and hierarchical modelling into a unified framework, so that all the model parameters can be learned toward the optimization of the objective function. We validate the effectiveness of the proposed model with two real datasets, and demonstrate its advantage over the state-of-the-art recommendation models with extensive experiments., Comment: 10 pages, 16 figures, The 28th ACM International Conference on Information and Knowledge Management (CIKM '19)

Accurate analysis of prostate cancer specific biomarkers plays an important role in the early diagnosis of prostate cancer. Traditional colorimetric lateral flow assay (LFA) has the limitations of low detection sensitivity and qualitative or semiquantitative detection. In this study, we developed a novel surface-enhanced Raman scattering (SERS)-based competitive LFA for the rapid and highly sensitive quantitative evaluation of prostate cancer antigen 3 (PCA3) mimic DNA. Herein, the competitive hybridization interaction with capture DNA between target PCA3 mimic DNA and reporter DNA-labeled SERS nanotags results in a change in the amount of SERS nanotags on the test line. The quantitative analysis of target PCA3 mimic DNA was realized by monitoring the Raman peak intensity of SERS nanotags on the test line. The limit of detection of PCA3 mimic DNA was estimated to be 3 fM, which is about three orders of magnitude more sensitive than that of a commercially available kit. By combining the outstanding characteristics of the well-established SERS-based competitive strategy and LFA platform, our design has strong potential for the early diagnosis of prostate cancer and other diseases.

Aspect Terms Extraction (ATE) and Aspect Categories Detection (ACD) are two fundamental sub-tasks for aspect-based sentiment analysis. Most of the existing works mainly focus on the ATE task or the co-extraction of aspect terms and opinion words, while few attention are paid to the ACD task. In this work, we propose a joint model to seamlessly integrate the ATE and ACD tasks into a multi-task learning framework. Each of the tasks is based on multi-layer Convolutional Neural Networks (CNNs) for computing high-level word representations, and produces a task-specific and a task-share vector. The task-share vector of one task is used to propagate information to the other, and guides the counterpart task to align the informative textual features to produce the task-specific vectors. Finally, a fully-connected layer with a softmax/sigmoid function is applied to the task-specific vectors for the specific information extraction. The rationale underlying the proposed joint model is that, aspect terms and aspect categories are semantically related, and the information propagated between the two tasks can help to capture the semantic alignments between the aspect terms and categories, and produce informative task-specific vectors. Moreover, the ATE task models local semantics at each position of a sentence, while the ACD task extracts global features of the whole sentence. The mutual interactions between local and global features, therefore, can reciprocally capture informative textual features for the information extraction tasks. We validate the effectiveness of the proposed model on two widely used datasets, and show its advantage over the state-of-the-art baselines. We also investigate the effectiveness of the multi-task framework by comparing the proposed model with its variants. Further, we study the robustness of the proposed model by presenting the model performance with respect to different hyperparameters. Finally, we provide visualization examples to gain a better understanding of the advantages the multi-task learning scheme.

A new type of electric- or thermal-responsive multilayer device composed of SiO2 bilayer inverse opal (IOP) and chiral nematic liquid crystals (N*LCs) was developed. Bilayer IOP was fabricated by layer-by-layer assembly of polystyrene (PS) spheres with two different sizes and showed a reflectance in an extended range of the near-infrared region. Furthermore, the electrically or thermally tunable reflectance of the bilayer-IOP-N*LC device was investigated. The device exhibited the photonic bandgap (PBG) of the N*LC-IOP composite structure with the application of an electric field (voltage-on), while it presented the reflectance of N*LCs without an electric field (voltage-off) and the electrically-responsive behaviour could be reversibly switched. Besides, the device exhibited a gradient redshift of reflectance as temperature increased below the clearing point (TC) while it showed the PBG of the N*LC-IOP composite structure when the temperature was above TC.

An activity recognition system essentially processes raw sensor data and maps them into latent activity classes. Most of the previous systems are built with supervised learning techniques and pre-defined data sources, and result in static models. However, in realistic and dynamic environments, original data sources may fail and new data sources become available, a robust activity recognition system should be able to perform evolution automatically with dynamic sensor availability in dynamic environments. In this paper, we propose methods that automatically incorporate dynamically available data sources to adapt and refine the recognition system at run-time. The system is built upon ensemble classifiers which can automatically choose the features with the most discriminative power. Extensive experimental results with publicly available datasets demonstrate the effectiveness of our methods. HighlightsPropose an activity recognition framework to incorporate dynamically discovered sensors automatically.Propose a method to select the most informative samples for retraining.Propose a novel way of combining AdaBoost with HMM&CRF for temporal regularization.

Neighborhood-based recommenders are a major class of Collaborative Filtering models. The intuition is to exploit neighbors with similar preferences for bridging unseen user-item pairs and alleviating data sparseness, in other words, learn the sub-graph representation of each user in a user graph. Many existing works propose neural attention networks to aggregate neighbors and place higher weights on the specific subsets of users for recommendation. However, the neighborhood information is not necessarily always informative, and the noises in the neighborhood can negatively affect the model performance. To address this issue, we propose a novel neighborhood-based recommender, where a hybrid gated network is designed to automatically separate similar neighbors from dissimilar (noisy) ones, and aggregate those similar neighbors to comprise neighborhood representations. The confidence in the neighborhood is also addressed by putting higher weights on the neighborhood representations if we are confident with the neighborhood information, and vice versa. In addition, a user-neighbor component is proposed to explicitly regularize user-neighbor proximity in latent space. These two components are combined into a unified model to complement each other for the recommendation task. Extensive experiments on three public datasets demonstrate that the proposed model consistently outperforms the state-of-the-art neighborhood-based recommenders. Furthermore, we study different variants of the proposed model to justify the underlying intuition of the proposed hybrid gated network and user-neighbor modeling components.

Traditional recommender systems encounter several challenges such as data sparsity and unexplained recommendation. To address these challenges, many works propose to exploit semantic information from review data. However, these methods have two major limitations in terms of the way to model textual features and capture textual interaction. For textual modeling, they simply concatenate all the reviews of a user/item into a single review. However, feature extraction at word/phrase level can violate the meaning of the original reviews. As for textual interaction, they defer the interactions to the prediction layer, making them fail to capture complex correlations between users and items. To address those limitations, we propose a novel Hierarchical Text Interaction model (HTI) for rating prediction. In HTI, we propose to model low-level word semantics and high-level review representations hierarchically. The hierarchy allows us to exploit textual features at different granularities. To further capture complex user–item interactions, we propose to exploit semantic correlations between each user–item pair at different hierarchies. At word level, we propose an attention mechanism specialized to each user–item pair, and capture the important words for representing each review. At review level, we mutually propagate textual features between the user and item, and capture the informative reviews. The aggregated review representations are integrated into a collaborative filtering framework for rating prediction. Experiments on five real-world datasets demonstrate that HTI outperforms state-of-the-art models by a large margin. Further case studies provide a deep insight into HTI’s ability to capture semantic correlations at different levels of granularities for rating prediction.

Fast association aims to quickly associate a large number of stations with an access point in 802.11ah networks. Existing fast association mechanisms suffer from efficiency, fairness and robustness problems. In this paper, we propose the mechanism of Fast Association based on Speculating the nUmber of Stations (FASUS). FASUS adopts new methods for retransmission, thresholding and adaptive round selection, which greatly improve the association performance. Experiments show that FASUS is able to reduce the association time by 67.1% when compared with the Linear Increase Linear Decrease (LILD) method, which is one of the best-known mechanisms for 802.11ah networks. We formulate a mathematical model to analyze the association process and to determine the optimal number of stations per round. Furthermore, we propose two methods to address the inter- and intra-network interference in congested networks. To improve the robustness and fairness of the network, we propose new solutions to address two possible attacks: 1) the Denial-of-Service (DoS) attack, which can disable the whole network, and 2) the selfish node attack, which may unfairly allow the attackers to associate in a much faster way than normal stations.

Cardiac troponin I (cTnI) was considered as the “gold standard” for acute myocardial infarction (AMI) diagnosis owing to its superior cardiac specificity for cardiac damage and showing little or no changes in patients with a skeletal muscle disease or trauma. Herein, a new signal amplification surface-enhanced Raman scattering (SERS) platform was developed for recognition and detection of cTnI by using gold nanoparticles (AuNPs), graphene oxide (GO) and magnetic beads (MB). Here, antibody/Raman reporter labeled AuNP–functionalized GO were employed as both SERS nanotags and signal amplification carriers. Monoclonal antibody modified MB were applied as the capture probe and separation agents. In the presence of cTnI, sandwich type immunocomplexes, “capture probe/target/SERS nanotags”, were formed through antibody–antigen–antibody interactions. Due to the strong SERS enhancement ability of the designed GO/AuNP complexes and a high binding chance between cTnI and the GO/AuNP complexes, the proposed SERS-based immunoassay could selectively detect cTnI with a high sensitivity (detection limit of 5 pg mL−1) and a good linearity was obtained in a range of 0.01–1000 ng mL−1. In addition, this method was also successfully applied for detecting cTnI in serum substitute media with a similar linear range. Furthermore, this strategy can be constructed with different kinds of antibodies and Raman reporters, and thus can be easily used for simultaneous detection of multiple biomarkers. Therefore, this proposed SERS-based signal amplification technique shows strong potential for the clinical diagnosis of AMI disease.

As the sheer volume of available micro-videos often undermines the users’ capability to choose the micro-videos, in this paper, we propose a multi-source multi-net micro-video recommendation model that recommends micro-videos fitting users’ best interests. Different from existing works, as micro-video inherits the characteristics of social platforms, we simultaneously incorporate multi-source content data of items and multi-networks of users to learn user and item representations for recommendation. This information can be complementary to each other in a way that multi-modality data can bridge the semantic gap among items, while multi-type user networks, such as following and reposting, are able to propagate the preferences among users. Furthermore, to discover the hidden categories of micro-videos that properly match users’ interests, we interactively learn the user-item representations. The resulted categorical representations are interacted with user representations to model user preferences at different level of hierarchies. Finally, multi-source content item data, multi-type user networks and hidden item categories are jointly modelled in a unified recommender, and the parameters of the model are collaboratively learned to boost the recommendation performance. Experiments on a real dataset demonstrate the effectiveness of the proposed model and its advantage over the state-of-the-art baselines.

In wireless 802.11 networks, MAC-layer rate control mechanisms such as Minstrel have an important role in selecting an appropriate transmission rate to maximize throughput and/or to minimize frame loss. Transmitting at a higher rate in good channel quality achieves high throughput, but it achieves a low or zero throughput if the channel quality is bad. In many MAC-layer rate control mechanisms, the selection of transmission rates depends on link quality metrics calculated based on MAC-layer acknowledgements (ACKs) of the successful transmissions of data frames. We found that injecting forged ACKs to acknowledge lost data frames can mislead most rate control mechanisms to select an unsustainable transmission rate resulting in frame losses. Due to the lack of protection of ACKs, the sender cannot detect and avoid this attack. In this paper, we develop a MAC-layer ACK spoofing mechanism and design experiments to exploit such a vulnerability in 802.11 networks. We analyze this vulnerability in different attack models for well-known MAC-layer rate control mechanisms. Our experiment results show that the spoofing attack can reduce the throughput to zero by tricking the sender to believe the highest rate should be used even when the channel quality is bad. To address this issue, we present a defense mechanism that is effective and light-weight. Experiment results confirm the effectiveness of the proposed solution with low communication overhead.

Organic second-order nonlinear optical materials are of great significance in the field of integrated photonics, optical data storage and optical information processing. In this work, two iodide salts with benzo[e]indol cations as electron-withdrawing groups were designed and synthesized, which are (E)-2-(4-hydroxy-3-methoxystyryl)-1,1,3-trimethyl-1H-benzo[e]indol-3-ium iodide (1-TB) and (E)-2-(3,4-dihydroxystyryl)-1,1,3-trimethyl-1H-benzo[e]indol-3-ium iodide (2-TB). Under the irradiation of 1064 nm laser, the second harmonic generations of the 1-TB and 2-TB powders are 760 and 580 times that of urea, respectively. Among them, crystals of 1-TB are successfully grown by a slow cooling method without using seed crystals. Bulk crystal with a size of 8.0 &times, 3.0 &times, 1.0 mm3 and many submillimeter single crystals with excellent optical quality are obtained.

We have demonstrated a facile and low-cost approach for the fabrication of binary “island” shaped arrays (BISA) with high-density hot spots as reproducible surface-enhanced Raman scattering (SERS) substrates by depositing a self-assembled monolayer Au nanoparticle (AuNP) film with small gaps onto a two-dimensional (2D) silica microsphere opal structure. By varying the size of silica spheres, the SERS performance of the BISA substrate with an enhancement factor (EF) of 3.74 × 1010 magnitude and the corresponding signal intensity deviation of below 8% using 770 nm silica sphere arrays were achieved. Compared with the assembled monolayer AuNP film on a planar substrate, the BISA enabled the installation of more AuNPs as a source of hot spots due to the undulation of morphology on the nanoscale within the designated laser-illumination area. In addition, a finite-difference time-domain (FDTD) simulation suggested that the BISA structure provided geometric conditions for increasing the intensity of the formed hot spots, and the strong periodic electric fields on the BISA are located not only in the gap between adjacent AuNPs, but also along the boundary of the neighboring island of silica spheres. Surface plasmon-decayed hot carriers (hot electrons and hot holes) from AuNPs can be applied in the field of energy conversion (i.e., photocatalysis), integrated with the SERS as a sensitive optical indicator to accurately monitor the catalytic reaction process. Furthermore, we examined the catalytic reaction process of the dimerization of 4-ATP into DMAB and found that photocatalytic activity could be tuned by changing the size of silica spheres. This study provides a new design route for the fabrication of the SERS platform with high sensitivity and reproducibility to detect molecules or improve catalyst efficiency.

Numerous methods have been proposed to address different aspects of human activity recognition. However, most of the previous approaches are static in terms of the data sources used for the recognition task. As sensors can be added or can fail and be replaced by different types of sensors, creating an activity recognition model that is able to leverage dynamically available sensors becomes important. In this paper, we propose methods for activity learning and activity recognition adaptation in environments with dynamic sensor deployments. Specifically, we propose sensor and activity context models to address the problem of sensor heterogeneity, so that sensor readings can be pre-processed and populated into the recognition system properly. Based on those context models, we propose the learning-to-rank method for activity learning and its adaptation. To model the temporal characteristics of the human behaviours, we add temporal regularization into the learning and prediction phases. We use comprehensive datasets to demonstrate effectiveness of the proposed method, and show its advantage over the conventional machine learning algorithms in terms of recognition accuracy. Our method outperforms hybrid models that combine typical machine learning methods with graphical models (i.e. HMM, CRF) for temporal smoothing.