Source Stream Research Articles

Assessing the energy resources policy agenda: Evidence from China's green express policy

Green Economy Recovery Plan is a respond to “wicked problems” in urban development. Exploration of the policy process and knowledge patterns in this plan can significantly enhance the science of urban decision-making. The development of electronic commerce has led to an increase in express waste, and the development of green consumption has become a focal point for the entire society. China's Green Transformation Express Packaging Policy was quickly adopted and disseminated in various cities; however, strategies for its promotion are inconsistent across regions. The multiple streams framework (MSF) is a mainstream approach that explores policy agenda setting. Based on the MSF, we employed a mixed methods research to examine the source streams factors that influence the agenda setting of the green transformation express packaging. First, we examine the negative effects of problem streams, policy streams, and political streams on the delay in policy agenda using data from the Chinese prefecture-level cities. Second, we confirm that policy entrepreneurs can advance policy agenda setting in partially coupled streams. Further, we uncovered different categories of policy entrepreneurs through qualitative interviews. This study fills a gap in the quantitative research on green policies within the Chinese context of the MSF and supports the role of partial coupling which is less applied in the MSF. The policy implications of this study are as follows: to consider an incentive mechanism for actively raising green fiscal demands can be included in the appraisal of local officials, to implement green recovery programs when most conditions are present, and to strengthen the role of experts and members of the Chinese People's Political Consultative Conference in green express policies.

Dynamics of Ca and Mg storage of non-woody debris in a subtropical forest headwater stream during the rainy season.

Forest headwater streams are the monumental cement for relating habitats of the terrene and water. Nutri-ent dynamics of non-woody debris in stream can directly and indirectly regulate the cycle and transport of forest nutrients, for example, Ca and Mg. In the rainy season (from March to August) of 2021, we monitored the dyna-mics of Ca and Mg storage of non-woody debris in a typical headwater stream in a subtropical forest. The results showed that total Ca and Mg storage of non-woody debris per unit area of stream ranged from 178.1 to 890.5 mg·m-2 and 13.8 to 61.6 mg·m-2 during the rainy season, respectively. The Ca and Mg storages of non-woody debris per unit area of stream during the rainy season displayed a pattern of first increase and then decrease, and overall a decrease pattern. The storage varied significantly among different sites, with higher values in stream source than others. The total Ca and Mg storage of non-woody debris positively correlated with precipitation, but negatively with stream water alkalinity, temperature, and dissolved oxygen. The variation of riparian forest type (e.g., Castano-psis carlesii forests or mixed coniferous forests) and with or without tributaries did not affect the storage of Ca and Mg in stream non-woody debris. During the rainy season, total Ca and Mg storage of non-woody debris in the headwater stream from forest generally decreased over time, which was mainly controlled by the characteristics of rainfall and stream.

Estimating the Routing Parameter of the Xin’anjiang Hydrological Model Based on Remote Sensing Data and Machine Learning

The parameters of hydrological models should be determined before applying those models to estimate or predict hydrological processes. The Xin’anjiang (XAJ) hydrological model is widely used throughout China. Since the prediction in ungauged basins (PUB) era, the regionalization of the XAJ model parameters has been a subject of intense focus; nevertheless, while many efforts have targeted parameters related to runoff yield using in-site data sets, classic regression has predominantly been applied. In this paper, we employed remotely sensed underlying surface data and a machine learning approach to establish models for estimating the runoff routing parameter, namely, CS, of the XAJ model. The study was conducted on 114 catchments from the Catchment Attributes and MEteorology for Large-sample Studies (CAMELS) data set, and the relationships between CS and various underlying surface characteristics were explored by a gradient-boosted regression tree (GBRT). The results showed that the drainage density, stream source density and area of the catchment were the three major factors with the most significant impact on CS. The best correlation coefficient (r), root mean square error (RMSE) and mean absolute error (MAE) between the GBRT-estimated and calibrated CS were 0.96, 0.06 and 0.04, respectively, verifying the good performance of GBRT in estimating CS. Although bias was noted between the GBRT-estimated and calibrated CS, runoff simulations using the GBRT-estimated CS could still achieve results comparable to those using the calibrated CS. Further validations based on two catchments in China confirmed the overall robustness and accuracy of simulating runoff processes using the GBRT-estimated CS. Our results confirm the following hypotheses: (1) with the help of large sample of catchments and associated remote sensing data, the ML-based approach can capture the nonstationary and nonlinear relationships between CS and the underlying surface characteristics and (2) CS estimated by ML from large samples has a robustness that can guarantee the overall performance of the XAJ mode. This study advances the methodology for quantitatively estimating the XAJ model parameters and can be extended to other parameters or other models.

Integrating Heterogeneous Stream and Historical Data Sources using SQL

Applications capable of integrating data from historical and streaming sources can make the most contextualized and enriched decision-making. However, the complexity of data integration over heterogeneous data sources can be a hard task for querying in this context. Approaches that facilitate data integration, abstracting details and formats of the primary sources can meet these needs. This work presents a framework that allows the integration of streaming and historical data in real-time, abstracting syntactic aspects of queries through the use of SQL as a standard language for querying heterogeneous sources. The framework was evaluated through an experiment using relational datasets and real data produced by sensors. The results point to the feasibility of the approach.

Sources and mean transit times of stream water in an intermittent river system: the upper Wimmera River, southeast Australia

Abstract. Determining the mean transit times (MTTs) and water sources in catchments at different flow conditions helps better understand river functioning, and manage river health and water resources. Despite being common in a range of environments, the MTTs and water sources in intermittent streams are much less well understood compared to perennial streams. Major ion geochemistry, stable isotopes, 14C, and 3H were used in this study to identify water sources and MTTs of the periodically intermittent upper Wimmera River from southeast Australia at different flow conditions, including zero-flow periods. The disconnected pool waters during the zero-flow period in the summer months of 2019 had 3H activities of 0.64 to 3.29 TU. These and the variations in total dissolved solids and stable isotopes imply that these pools contained a mixture of older groundwater and younger stream water impacted by evaporation. 3H activities during the high-flow period in July 2019 were 1.85 to 3.00 TU, yielding MTTs of up to 17 years. The 3H activities at moderate and low-flow conditions in September and November 2019 ranged from 2.26 to 2.88 TU, implying MTTs of 1.6 to 7.8 years. Regional groundwater near the Wimmera River had 3H activities of < 0.02 to 0.45 TU and 14C activities of 57 to 103 pMC, and was not recharged by the river at high flows. The Wimmera River and other intermittent streams in southeast Australia are sustained by younger catchment waters from relatively small near-river stores than comparable perennial streams, which have older deeper regional groundwater inputs. This results in these intermittent streams being more susceptible to short-term changes in climate and necessitates the protection of near-river corridors to maintain the health of the riverine systems.

Issues and Strategies of Localising Sensitive Audiovisual Elements in Game Streaming: A Case Study on Overwatch League (OWL) Chinese Streaming

This paper is a participant-observer case study of the Overwatch League (OWL) 2021 Chinese streaming. Within the contemporary dynamic and diversified mediascape of China, esports streaming has become increasingly popular (Taylor, 2018; Wohn & Freeman, 2020). Esports streaming allows spectators to watch esports players’ intense competitions in real time and interact through danmu (bullet comments), gift sending systems, or via streaming lotteries (Lee et al., 2019; Abarbanel & Johnson, 2020; Zhang & Cassany, 2020). The Chinese streaming of Overwatch League (OWL), a series of Overwatch (OW) esports tournaments owned by Activision Blizzard, on Bilibili can be considered a game streaming localisation practice which localises audiovisual content from the original English-language streaming source into Chinese and then innovates upon it in the form of a number of local streaming programmes. The OWL Chinese streaming has disclosed many issues and strategies taken toward sensitive elements like game slang, cultural differences, the presence of erotic or violent content, gambling, media accessibility and censorship laws. The analysis of the selected cases in this article extends the current theories dealing with game localisation (e.g. Zhang Xiaochun, 2012; O’Hagan & Mangiron, 2013; Mangiron et al., 2014) into a new layer, dealing with game streaming localisation. The findings suggest the complications of game streaming content, including the controversial impacts of localisation strategies on sensitive elements, cross-media marketing of streaming practices and the participation of various stakeholders (e.g. officials, sponsors, fans) in game streaming localisation.

Research on Multimodal Summarization Based on Sequence to Sequence Model

With the rapid development of information technology in the 21st century, the data on the Internet is growing exponentially, which makes it difficult for users to receive useful information from an endless stream of information sources. Therefore, as a kind of data compression task, automatic summarization technology has gradually become a research hotspot in academia and industry. This paper mainly introduces the multimodal summarization method from sequence to sequence model, then introduces the current multimodal summarization evaluation methods and finally gives some conclusions and thoughts.

Modelling stream nutrient source and yield by coupling hydrological processes and nutrient budgets from an experimental headwater catchment with orchard expansion in the Changjiang River Delta area

AbstractAlthough watershed nutrient exports have been well studied and modelled, the dynamic changes in the contributions of individual nutrient sources to reservoirs are less certain under changing land use and climate. The intensification of agriculture has led to expansion in plantation areas and the corresponding increased use of chemical and organic fertilisers in many headwater catchments in recent decades in China. Nutrient yields are potentially substantial because of the expansion of agricultural ecosystems. Here, in combination with biogeochemical nutrient budgets, we developed a synthesis model to quantify daily nutrient yields in terms of dissolved inorganic nitrogen (DIN), dissolved organic nitrogen (DON), dissolved inorganic phosphorus (DIP) and dissolved organic phosphorus (DOP). We also identified their sources in an experimental headwater reservoir catchment based on high‐frequency observations of daily precipitation with nutrient concentrations in the stream outlets from January 2020 to December 2021. Annual DIN, DON, DIP and DOP yields were 31.68 ± 2.38, 5.98 ± 0.41, 1.12 ± 0.08 and 0.37 ± 0.27 kg ha−1 yr−1, respectively. All nutrient yields showed significant seasonal patterns, with over 60% of the total nutrient yield obtained from May to July. Our model suggests that over 70% of the annual nutrient yields were driven by rainfall exceeding 40 mm. Approximately 70% of catchment nutrient yields were from chemical and organic fertiliser applications. Triple nitrate isotopes (δ15N‐NO3−, δ18O‐NO3− and δ17O‐NO3−) revealed that stream nitrate sources and their contributions were almost consistent with the estimation from our model. The estimates demonstrated that orchard expansion by 25% of the total area led to a 63% increase in dissolved total nitrogen and an 88% increase in dissolved total phosphorus exports. Our results reveal that diffuse sources from orchard plantations are an important yet overlooked source of nutrients to the receiving headwater reservoir budget.

ACDC: Online unsupervised cross-domain adaptation

We consider the problem of online unsupervised cross-domain adaptation, where two independent but related data streams with different feature spaces – a fully labeled source stream and an unlabeled target stream – are learned together. Unique characteristics and challenges such as covariate shift, asynchronous concept drifts, and contrasting data throughput arise. We propose ACDC, an adversarial unsupervised domain adaptation framework that handles multiple data streams with a complete self-evolving neural network structure that reacts to these defiances. ACDC encapsulates three modules into a single model: A denoising autoencoder that extracts features, an adversarial module that performs domain conversion, and an estimator that learns the source stream and predicts the target stream. ACDC is a flexible and expandable framework with little hyper-parameter tunability. Our experimental results under the prequential test-then-train protocol indicate an improvement in target accuracy over the baseline methods, achieving more than a 10% increase in some cases.

Assessment of selected pesticide contamination of surface and ground water samples from an agrarian settlement in Kuje area council, Abuja, Nigeria

This study aimed at evaluating the presence and levels of some selected pesticides in both surface and ground water in an agrarian settlement in Kuje area council, Abuja, Nigeria. Samples of water were collected from three areas in Kuje, Abuja, Nigeria namely: Gaube, Kasada and Pasali. The water samples were obtained from borehole, well and stream sources in triplicate. Dissolved oxygen, electrical conductivity, pH, total dissolved solids, and turbidity were the physicochemical parameters determined in the water samples. Five (5) pesticides namely: Lambda-cyhalothrin, Chlorpyrifos, Glyphosate, Dichlorvos and Cypermethrin were investigated using highperformance liquid chromatography-diode array detector (HPLC-DAD). The findings from the study show that pH values were between 5.70±0.30 and 7.82±0.12, electrical conductivity ranged from 61.31±38.41 to 242.81±1.40 μS/cm, dissolved oxygen 5.58±0.69 to 6.40±2.15 mg/L and turbidity values ranged from 1.42±0.51 to 143.11±11.03 NTU. Lambda-cyhalothrin levels were from 0.06±0.00 to 80.30±138.31μg/L, Chlorpyrifos ranged from 139.99±69.07 to 243.49±197.94 μg/L, the concentrations of Dichlorvos were found to be between 86.65±667.12 and 131.22±3.20 μg/L, Glyphosate and Cypermethrin were not detected in any of the samples tested. Lambda-cyhalothrin correlated positively with Dichlorvos (0.721) and Chlorpyrifos (0.951) in stream water sample from Pasali at p<0.05. The concentrations of the three pesticides exceeded the permissible limit except for the Lambda-cyhalothrin in stream water sample from Pasali and this calls for worry because of the potential adverse effects these could cause the residents of Kuje that continuously use and consume water from these sources.

Instance exploitation for learning temporary concepts from sparsely labeled drifting data streams

Continual learning from streaming data sources becomes more and more popular due to the increasing number of online tools and systems. Dealing with dynamic and everlasting problems poses new challenges for which traditional batch-based offline algorithms turn out to be insufficient in terms of computational time and predictive performance. One of the most crucial limitations is that we cannot assume having an access to a finite and complete data set – we always have to be ready for new data that may complement our model. This poses a critical problem of providing labels for potentially unbounded streams. In real world, we are forced to deal with very strict budget limitations, therefore, we will most likely face the scarcity of annotated instances, which are essential in supervised learning. In our work, we emphasize this problem and propose a novel instance exploitation technique. We show that when: (i) data is characterized by temporary non-stationary concepts, and (ii) there are very few labels spanned across a long time horizon, it is actually better to risk overfitting and adapt models more aggressively by exploiting the only labeled instances we have, instead of sticking to a standard learning mode and suffering from severe underfitting. We present different strategies and configurations for our methods, as well as an ensemble algorithm that attempts to maintain a sweet spot between risky and normal adaptation. Finally, we conduct a complex in-depth comparative analysis of our methods, using state-of-the-art streaming algorithms relevant for the given problem.

Propagation, Growth and Haematology of Fish Cultured in Water from a Natural Stream and Underground Tubewell Sources

Propagation, growth and haematology of Clarias gariepinus bred and reared in water from a natural stream at 5o28’N, 7o32’E (Nature stream) and two wells at 5.4810N, 7.5370E (well-1) and 5.4770N, 7.5390E (well-2) were studied. The study was carried out in search of suitable water within close proximity for fish propagation. Growth was evaluated once in two weeks. Haematological parameters of the fingerlings were examined at 70 day post hatch. There was significant difference (P<0.05) in egg fertilization, hatchability, larval survival, growth and blood parameters among the treatments. Fish exposed to water from the natural stream impacted best propagation indices including fastest growth and highest values of Packed Cell Volume (PCV) and Mean Corpuscular Haemoglobin Concentration (MCHC). Fish for well-1 and well-2 were not statistically different (P>0.05) in PCV and MCHC. Apart from these two parameters, values of all the other blood parameters including growth were significantly lower (P<0.05) in fish for well-2 than those for well-1. The disparities observed signifies differences in water quality which is indicative for the presence of contaminant not identified in this study. The report recommended inclusion of heavy metals and organic mineral analysis in future study to identify the cause of the observed alterations.

Seasonal and functional variation in the trophic base of intermittent Alpine streams

AbstractIn high‐altitude Alpine streams, seasonal cycles of snowmelt, glacial melt, and rainfall drive variation in the availability of algal food resources. Yet, high‐altitude streams also exhibit varying degrees of flow intermittency, ranging from solely winter‐drying streams to others that dry periodically throughout summer and autumn. These environmental drivers may interact in different ways to determine the functional trophic base of macroinvertebrates inhabiting high‐altitude streams. Here, we estimated the proportional contribution of autochthonous resources to the assimilated diets of benthic macroinvertebrates in 26 headwater streams of Val Roseg, a high Alpine glacial catchment, using stable isotope analysis (δ13C and δ15N) of different macroinvertebrate families and their potential food sources. We compared dietary estimates along a gradient of flow intermittency and across three seasons (Alpine spring, summer, and autumn). Assimilation from autochthonous sources was highest for collector‐gatherers and filter feeders in spring, and for grazers in summer. Grazers had higher estimated assimilation from autochthonous sources in intermittent streams than in perennial streams, particularly in summer, while collector‐gatherers showed little effect of flow intermittency on dietary estimates. However, responses were highly taxon‐specific, with different responses to variation in flow intermittency and season across families within functional groups. Our results suggest that frequent summer‐drying events represent tradeoffs between greater access to algal food resources and a higher risk of desiccation, but that differing life history and functional feeding traits across macroinvertebrate taxa drive marked variation in the risks or benefits associated with inhabitants of drying streams.

On Differentiating Multiple Types of ULF Magnetospheric Waves in Response to Solar Wind Periodic Density Structures.

Identifying the nature and source of ultra‐low frequencies (ULF) waves (f ⪅ 4 mHz) at discrete frequencies in the Earth's magnetosphere is a complex task. The challenge comes from the simultaneous occurrence of externally and internally generated waves, and the ability to robustly identify such perturbations. Using a recently developed robust spectral analysis procedure, we study an interval that exhibited in magnetic field measurements at geosynchronous orbit and in‐ground magnetic observatories both internally supported and externally generated ULF waves. The event occurred on 9 November 2002 during the interaction of the magnetosphere with two interplanetary shocks that were followed by a train of 90 min solar wind periodic density structures. Using the Wang‐Sheeley‐Arge model, we mapped the source of this solar wind stream to an active region and a mid‐latitude coronal hole just prior to crossing the Heliospheric current sheet. In both the solar wind density and magnetospheric field fluctuations, we separated broad power increases from enhancements at specific frequencies. For the waves at discrete frequencies, we used the combination of satellite and ground magnetometer observations to identify differences in frequency, polarization, and observed magnetospheric locations. The magnetospheric response was characterized by: (a) forced breathing by periodic solar wind dynamic pressure variations below ≈1 mHz, (b) a combination of directly driven oscillations and wave modes triggered by additional mechanisms (e.g., shock and interplanetary magnetic field discontinuity impact, and substorm activity) between ≈1 and 4 mHz, and (c) largely triggered modes above ≈4 mHz.

Predicting sediment connectivity in a mountain basin: A quantitative analysis of the index of connectivity

AbstractSediment connectivity underlies the relationships between sediment source areas and stream network and the use of topography‐based indices, such as the index of connectivity (IC), has become a common approach to investigate these relationships. However, few studies have tested the potential of the IC as an independent variable for predicting observable processes. In this study, the aim is the validation of the IC as a semi‐quantitative tool for depicting structural connectivity and predicting sediment dynamics linking sediment sources to the channel network after a large infrequent disturbance in the Rio Cordon catchment (Italy). First, the assessment of lateral connectivity was carried out using a combination of remote sensing data and field observations. In this way, it was possible to classify the sediment sources into disconnected, connected or functionally connected. Second, logistic regression analyses were performed using different IC variants, with different weighting factors and spatial resolutions as independent variables and connectivity status as dependent variable. Hence, the predictive capacity of the IC was tested and a threshold was derived to distinguish between low and high sediment connectivity. The results showed that only 15% of sediment sources are connected to the stream network. Moreover, the IC proved to be a reliable tool for depicting structural connectivity when using surface roughness as weighting factor and high‐resolution digital terrain models. On the contrary, it failed at predicting sediment dynamics as it did not classify correctly the functionally connected areas, so it cannot infer functional sediment connectivity in this case study. As a consequence, it was possible to obtain an IC threshold of −2.32, capable of differentiating between low and high structural sediment connectivity and useful for future management decisions. The threshold, in fact, represents an objective boundary between disconnected and connected sources in the Rio Cordon catchment.

Techniques for Skeletal-Based Animation in Massive Crowd Simulations

Crowd systems play an important role in virtual environment applications, such as those used in entertainment, education, training, and different simulation systems. Performance and scalability are key factors, and it is desirable for crowds to be simulated with as few resources as possible while providing variety and realism for agents. This paper focuses on improving the performance, variety, and usability of crowd animation systems. Performing the blending operation on the Graphics Processing Unit (GPU) side requires no additional memory other than the source and target animation streams and greatly increases the number of agents that can simultaneously transition from one state to another. A time dilation offset feature helps applications with a large number of animation assets and/or agents to achieve sufficient visual quality, variety, and good performance at the same time by moving animation streams between the running and paused states at runtime. Splitting agents into parts not only reduces asset creation costs by eliminating the need to create permutations of skeletons and assets but also allows users to attach parts dynamically to agents.

Resource scheduling and provisioning for processing of dynamic stream workflows under latency constraints

There has been an increasing demand for large-scale processing of stream workflows. Many streaming data sources, such as video and audio, differ in terms of the type and density of information they contain over time, which require dynamic workflow structures where paths change in response to data variations. Additionally, information is often only inferred from examining discrete chunks and hence, cannot be arbitrarily parallelised like other streaming data sources. Stream processes using these data sources often operate as dynamic workflows, where different paths are selected based on prior processing information, relating to differences in the nature of the data itself. To better enable low latency analysis in such scenarios, this work investigates strategies to schedule and provision streaming dynamic workflow scenarios of various complexities and degrees of unpredictability. It is found that, as scenarios become increasingly uncertain and unpredictable, it is better to use simpler strategies which do not make assumptions about workflow structures or completion times.

Machine learning optimization of a novel geothermal driven system with LNG heat sink for hydrogen production and liquefaction

Hydrogen production and liquefaction based on geothermal energy is a potential route for the future hydrogen economy. In the current study, a novel integrated system with power generation and cooling capabilities is designed which uses geothermal energy as a heat source and LNG stream as a heat sink. All the generated power by the system is delivered to the PEM electrolyzer to produce hydrogen and liquefied it through a Claude cycle. A comprehensive investigation is carried out to evaluate the performance of the system from a thermodynamic and economic points of view. The analysis shows that the hydrogen production rate is 106.8 kg/h if all the electricity is delivered to PEM electrolyzer. Also, PEM electrolyzer with 93.92 $/h and LNG vaporizer with 5.43 MW have the foremostimpact on total cost rate and exergy destruction, respectively. Moreover, a parametric study is performed to understand the effects of input parameters on the performance of the system. In order to optimize hydrogen production rate, total cost rate, and exergy efficiency of the system, a multi-objective optimization process is applied to the system by coupling the artificial neural network with the genetic algorithm. From the optimization procedure, the optimum values ​​of hydrogen production rate, total cost rate, and exergy efficiency are obtained as 154.95 (kg/h), 291.36 ($/h), 23.34%, respectively. At these conditions, cooling capacity and levelized cost of hydrogen are 5.25 MW and 1.827 $/ kg, correspondingly.

RealTE: Real-time Trajectory Estimation Over Large-Scale Road Networks

Real-time traffic estimation focuses on predicting the travel time of one travel path, which is capable of helping drivers selecting an appropriate or favor path. Statistical analysis or neural network approaches have been explored to predict the travel time on a massive volume of traffic data. These methods need to be updated when the traffic varies frequently, which incurs tremendous overhead. We build a system RealTER⁢e⁢a⁢l⁢T⁢E, implemented on a popular and open source streaming system StormS⁢t⁢o⁢r⁢m to quickly deal with high speed trajectory data. In RealTER⁢e⁢a⁢l⁢T⁢E, we propose a locality-sensitive partition and deployment algorithm for a large road network. A histogram estimation approach is adopted to predict the traffic. This approach is general and able to be incremental updated in parallel. Extensive experiments are conducted on six real road networks and the results illustrate RealTE achieves higher throughput and lower prediction error than existing methods. The runtime of a traffic estimation is less than 11 seconds over a large road network and it takes only 619619 microseconds for model updates.

A new multi-stream approach using acoustic and visual features for robust speech recognition system

Building a robust Automatic Speech Recognition (ASR) system and improving recognition accuracy in adverse conditions is still a challenging task. One way to improve the robustness of an ASR system is combining information from multiple sources (streams). A multi-stream approach which handles the multiple inputs at the model level is the key contribution of our work. Standard mic (Sm), Throat mic (Tm), and Lip reading (Lr) are the various source streams that have been used. This work explores a static weighted two stream HMM (TSH) and multi-stream HMM (MSH) model for the bimodal and multimodal systems. A syllabic units of the Hindi language database categorized into three categories – Vowel, Place of Articulation (POA), and Manner of Articulation (MOA) are used for training and testing. In this study, four types of TSH are proposed for the combination of bimodal ((Sm+Tm), (Tm+Lr), (Sm+Lr), (Lm+Lm)) systems and one type MSH is proposed for multimodal (Sm+Tm+Lr) system in a synchronous and asynchronous manner. Mel Frequency Cepstral Coefficient (MFCC) features are used for Sm and Tm signals. Combined pixel-motion based features (DCT/DWT-MHI) are used for Lr signals. Among these two features, DWT outperforms than DCT and used as a feature for visual speech. Experiments were conducted for bimodal and multimodal system. The proposed MSH approach shows improvements of 1.36%, 6.21%, and 5.8% in recognition accuracy for Vowel, POA and MOA category respectively, as compared to bimodal systems.