74 resultados para human activity recognition
Resumo:
This paper presents a method for rational behaviour recognition that combines vision-based pose estimation with knowledge modeling and reasoning. The proposed method consists of two stages. First, RGB-D images are used in the estimation of the body postures. Then, estimated actions are evaluated to verify that they make sense. This method requires rational behaviour to be exhibited. To comply with this requirement, this work proposes a rational RGB-D dataset with two types of sequences, some for training and some for testing. Preliminary results show the addition of knowledge modeling and reasoning leads to a significant increase of recognition accuracy when compared to a system based only on computer vision.
Resumo:
Support vector machines (SVMs), though accurate, are not preferred in applications requiring high classification speed or when deployed in systems of limited computational resources, due to the large number of support vectors involved in the model. To overcome this problem we have devised a primal SVM method with the following properties: (1) it solves for the SVM representation without the need to invoke the representer theorem, (2) forward and backward selections are combined to approach the final globally optimal solution, and (3) a criterion is introduced for identification of support vectors leading to a much reduced support vector set. In addition to introducing this method the paper analyzes the complexity of the algorithm and presents test results on three public benchmark problems and a human activity recognition application. These applications demonstrate the effectiveness and efficiency of the proposed algorithm.
--------------------------------------------------------------------------------
Resumo:
This research presents a fast algorithm for projected support vector machines (PSVM) by selecting a basis vector set (BVS) for the kernel-induced feature space, the training points are projected onto the subspace spanned by the selected BVS. A standard linear support vector machine (SVM) is then produced in the subspace with the projected training points. As the dimension of the subspace is determined by the size of the selected basis vector set, the size of the produced SVM expansion can be specified. A two-stage algorithm is derived which selects and refines the basis vector set achieving a locally optimal model. The model expansion coefficients and bias are updated recursively for increase and decrease in the basis set and support vector set. The condition for a point to be classed as outside the current basis vector and selected as a new basis vector is derived and embedded in the recursive procedure. This guarantees the linear independence of the produced basis set. The proposed algorithm is tested and compared with an existing sparse primal SVM (SpSVM) and a standard SVM (LibSVM) on seven public benchmark classification problems. Our new algorithm is designed for use in the application area of human activity recognition using smart devices and embedded sensors where their sometimes limited memory and processing resources must be exploited to the full and the more robust and accurate the classification the more satisfied the user. Experimental results demonstrate the effectiveness and efficiency of the proposed algorithm. This work builds upon a previously published algorithm specifically created for activity recognition within mobile applications for the EU Haptimap project [1]. The algorithms detailed in this paper are more memory and resource efficient making them suitable for use with bigger data sets and more easily trained SVMs.
Resumo:
Purpose
– Information science has been conceptualized as a partly unreflexive response to developments in information and computer technology, and, most powerfully, as part of the gestalt of the computer. The computer was viewed as an historical accident in the original formulation of the gestalt. An alternative, and timely, approach to understanding, and then dissolving, the gestalt would be to address the motivating technology directly, fully recognizing it as a radical human construction. This paper aims to address the issues.
Design/methodology/approach
– The paper adopts a social epistemological perspective and is concerned with collective, rather than primarily individual, ways of knowing.
Findings
– Information technology tends to be received as objectively given, autonomously developing, and causing but not itself caused, by the language of discussions in information science. It has also been characterized as artificial, in the sense of unnatural, and sometimes as threatening. Attitudes to technology are implied, rather than explicit, and can appear weak when articulated, corresponding to collective repression.
Research limitations/implications
– Receiving technology as objectively given has an analogy with the Platonist view of mathematical propositions as discovered, in its exclusion of human activity, opening up the possibility of a comparable critique which insists on human agency.
Originality/value
– Apprehensions of information technology have been raised to consciousness, exposing their limitations.
Resumo:
There has been considerable uncertainty about the nature of Pleistocene environments colonised by the first modern humans in Island SE Asia, and about the vegetation of the Last Glacial Maximum (LGM) in the region. Here, the palynology from a series of exposures in the Great Cave of Niah, Sarawak, Malaysian Borneo, spanning a period from ca. 52,000 to 5000 BP is described. Vegetation during this period was climate-driven and often highly unstable. Interstadials are marked by lowland forest, sometimes rather dry and at times by mangroves. Stadials are indicated by taxa characteristic of open environments or, as at the LGM, by highly disturbed rather open forest. Stadials are also characterised by taxa now restricted to 1000-1600 m above sea level, suggesting temperature declines of ca 7-9 C relative to present, by comparison with modern lapse rates. The practice of biomass burning appears associated with the earliest human activity in the cave.
Resumo:
Human action recognition is an important problem in computer vision, which has been applied to many applications. However, how to learn an accurate and discriminative representation of videos based on the features extracted from videos still remains to be a challenging problem. In this paper, we propose a novel method named low-rank representation based action recognition to recognize human actions. Given a dictionary, low-rank representation aims at finding the lowestrank representation of all data, which can capture the global data structures. According to its characteristics, low-rank representation is robust against noises. Experimental results demonstrate the effectiveness of the proposed approach on several publicly available datasets.
Resumo:
Recent research in Europe, Africa, and Southeast Asia suggests that we can no longer assume a direct and exclusive link between anatomically modern humans and behavioral modernity (the 'human revolution'), and assume that the presence of either one implies the presence of the other: discussions of the emergence of cultural complexity have to proceed with greater scrutiny of the evidence on a site-by-site basis to establish secure associations between the archaeology present there and the hominins who created it. This paper presents one such case study: Niah Cave in Sarawak on the island of Borneo, famous for the discovery in 1958 in the West Mouth of the Great Cave of a modern human skull, the 'Deep Skull,' controversially associated with radiocarbon dates of ca. 40,000 years before the present. A new chronostratigraphy has been developed through a re-investigation of the lithostratigraphy left by the earlier excavations, AMS-dating using three different comparative pre-treatments including ABOX of charcoal, and U-series using the Diffusion-Absorption model applied to fragments of bones from the Deep Skull itself. Stratigraphic reasons for earlier uncertainties about the antiquity of the skull are examined, and it is shown not to be an `intrusive' artifact. It was probably excavated from fluvial-pond-desiccation deposits that accumulated episodically in a shallow basin immediately behind the cave entrance lip, in a climate that ranged from times of comparative aridity with complete desiccation, to episodes of greater surface wetness, changes attributed to regional climatic fluctuations. Vegetation outside the cave varied significantly over time, including wet lowland forest, montane forest, savannah, and grassland. The new dates and the lithostratigraphy relate the Deep Skull to evidence of episodes of human activity that range in date from ca. 46,000 to ca. 34,000 years ago. Initial investigations of sediment scorching, pollen, palynomorphs, phytoliths, plant macrofossils, and starch grains recovered from existing exposures, and of vertebrates from the current and the earlier excavations, suggest that human foraging during these times was marked by habitat-tailored hunting technologies, the collection and processing of toxic plants for consumption, and, perhaps, the use of fire at some forest-edges. The Niah evidence demonstrates the sophisticated nature of the subsistence behavior developed by modern humans to exploit the tropical environments that they encountered in Southeast Asia, including rainforest. (c) 2006 Elsevier Ltd. All rights reserved.
Resumo:
Human activity has undoubtedly had a major impact on Holocene forested ecosystems, with the concurrent expansion of plants and animals associated with cleared landscapes and pasture, also known as 'culture-steppe'. However, this anthropogenic perspective may have underestimated the contribution of autogenic disturbance (e.g. wind-throw, fire), or a mixture of autogenic and anthropogenic processes, within early Holocene forests. Entomologists have long argued that the north European primary forest was probably similar in structure to pasture woodland. This idea has received support from the conservation biologist Frans Vera, who has recently strongly argued that the role of large herbivores in maintaining open forests in the primeval landscapes of Europe has been seriously underestimated. This paper reviews this debate from a fossil invertebrate perspective and looks at several early Holocene insect assemblages. Although wood taxa are indeed important during this period, species typical of open areas and grassland and dung beetles, usually associated with the dung of grazing animals, are persistent presences in many early woodland faunas. We also suggest that fire and other natural disturbance agents appear to have played an important ecological role in some of these forests, maintaining open areas and creating open vegetation islands within these systems. More work, however, is required to ascertain the role of grazing animals, but we conclude that fossil insects have a significant contribution to make to this debate. This evidence has fundamental implications in terms of how the palaeoecological record is interpreted, particularly by environmental archaeologists and palaeoecologists who may be more interested in identifying human-environment interactions rather than the ecological processes which may be preserved within palaeoecological records.
Resumo:
The extent to which North Atlantic Holocene climatic perturbations influenced past human societies is an area of considerable uncertainty and fierce debate. Ireland is ideally placed to help resolve this issue, being occupied for over 9000 yr and located on the eastern Atlantic seaboard, a region dominated by westerly airflow. Irish bog and lake tree populations provide unambiguous evidence of major shifts in surface moisture through the Holocene similar to cycles recorded in the marine realm of the North Atlantic, indicating significant changes in the latitude and intensity of zonal atmospheric circulation across the region. To test for human response to these cycles we summed the probabilities of 465 radiocarbon ages obtained from Irish archaeological contexts and observe enhanced archaeological visibility during periods of sustained wet conditions. These results suggest either increasing density of human populations in key, often defensive locations, and/or the development of subsistence strategies to overcome changing conditions, the latter recently proposed as a significant factor in avoiding societal collapse. Regardless, we demonstrate environmental change is a significantly more important factor in influencing human activity in the landscape than has hitherto been acknowledged.
Resumo:
During recent reinvestigations in the Great Cave of Niah in Borneo, the ‘Hell Trench’ sedimentary sequence seen by earlier excavators was re-exposed. Early excavations here yielded the earliest anatomically-modern human remains in island Southeast Asia. Calibrated radiocarbon dates, pollen, algal microfossils, palynofacies, granulometry and geochemistry of the ‘Hell Trench’ sequence provide information about environmental and vegetational changes, elements of geomorphic history and information about human activity. The ‘Hell’ sediments were laid down episodically in an ephemeral stream or pool. The pollen suggests cyclically changing vegetation with forest habitats alternating with more open environments; indicating that phases with both temperatures and precipitation reduced compared with the present. These events can be correlated with global climate change sequences to produce a provisional dating framework. During some forest phases, high counts of Justicia, a plant which today colonises recently burnt forest areas, point to fire in the landscape. This may be evidence for biomass burning by humans, presumably to maintain forest-edge habitats. There is evidence from palynofacies for fire on the cave floor in the ‘Hell’ area. Since the area sampled is beyond the limit of plant growth, this is evidence for human activity. The first such evidence is during an episode with significant grassland indicators, suggesting that people may have reached the site during a climatic phase characterised by relatively open habitats ~50 ka. Thereafter, people were able to maintain a relatively consistent presence at Niah. The human use of the ‘Hell’ area seems to have intensified through time, probably because changes in the local hydrological regime made the area dryer and more suitable for human use.
Resumo:
Palaeoecological methods can provide an environmental context for archaeological sites, enabling the nature of past human activity to be explored from an indirect but alternative perspective. Through a palynological study of a small fen wetland located within the catchment of a multi-period prehistoric complex at Ballynahatty, Co. Down, Northern Ireland, we reconstruct the vegetation history of the area during the early prehistoric period. The pollen record reveals tentative evidence for Mesolithic activity in the area at 6410-6220 cal. BC, with woodland disturbance identified during the Mesolithic-Neolithic transitional period ca. 4430-3890 cal. BC. A more significant impact on the landscape is observed in the Early Neolithic from 3944-3702 cal. BC, with an opening up of the forests and the establishment of a mixed agricultural economy. This activity precedes and continues to be evident during the Mid-Neolithic during which megalithic tombs and related burial sites were constructed at Ballynahatty. Due to chronological uncertainties and a possible hiatus in peat accumulation in the fen, the contemporary environment of the Ballynahatty timber circle complex (constructed and used ca. 3080-2490 cal. BC) and henge (dating to the third millennium cal. BC) cannot certainly be established. Nevertheless, the pollen record suggests that the landscape remained open through to the Bronze Age, implying a long continuity of human activity in the area. These findings support the idea that the Ballynahatty prehistoric complex was the product of a gradual and repeated restructuring of the ritual and ceremonial landscape whose significance continued to be recognised throughout the early prehistoric period.
Resumo:
In 2003, the remains of an Early Iron Age bog body, known as ‘Oldcroghan Man’, were recovered during the cutting of a drainage ditch in a bog in the Irish Midlands. Only some fingernails and a withe fragment remained undisturbed in situ in the drain face, providing the sole evidence for the original position of the body. A detailed reconstruction of the depositional context of the body has been undertaken through multi-proxy analyses of a peat monolith collected at the findspot. The palynological record shows that the surrounding area was the focus of intensive human activity during the Later Bronze Age, but was largely abandoned during the Bronze Age–Iron transition in the mid-first millennium BC. In the mid-4th century BC, a bog pool developed at the site, evidenced in the stratigraphic, plant macrofossil, testate amoebae and coleopteran records. Plant macrofossil and pollen analysis of peat samples associated with the fingernails suggests that the body was deposited in this pool most likely during the 3rd century BC. The absence of carrion beetle fauna points to complete submergence of the body within the pool. Deposition occurred shortly before or around the time that the surrounding area again became the focus of woodland clearance, as seen in the extended pollen record from the peat monolith. This period corresponds to the Early Iron Age in Ireland, during which renewed cultural connections with Britain and continental Europe can be seen in the archaeological record and widespread forest clearance is recorded in pollen records from across Ireland. The palaeoenvironmental results indicate, therefore, that the demise of Oldcroghan Man took place at a pivotal time of socio-economic and perhaps political change.
Resumo:
The recent article by Fenton (Fenton JH. 2008. A postulated natural origin for the open landscape of upland Scotland. Plant Ecology & Diversity 1:115–127) has argued that the landscapes of upland Scotland are treeless because of long-term deterioration of soil conditions. There are reasons for thinking that this might be the case in the absence of human activity. However, there have been considerable anthropogenic pressures on these landscapes for several millenia, documented archaeologically and palaeoecologically. Attempting to exclude these pressures from the discussion can only lead to an incomplete and misleading account of a complex series of changes involving an interaction which includes natural vegetational and environmental processes, climatic changes and human pressures.
Resumo:
Caves have yielded some of the most globally important archaeological sequences, but often their interpretation has suffered from assumptions about cave sedimentary processes. Caves contain distinctive sedimentary environments: this has major implications for the understanding of contained archaeological materials. This paper describes and analyses the Holocene sediments in the Haua Fteah, a sequence regarded as essentially continuous by the original excavator. 50 years after it was first excavated, the Haua’s Epipalaeolithic to post-Classical chronological range and rich finds make it still the key Holocene archaeological site in North Africa. The reassessment shows, however, that the sequence is strongly discontinuous and this has major implications for the reinterpretation of the site, as the highlyresolved archaeological record is thus likely to reflect a series of brief occupations, rather than continuous human activity. As with many caves, the sedimentary record in the Haua Fteah is an extremely sensitive indicator of environments and processes in the wider landscape. Secure understanding of sedimentary process, from analysis of the highly individual records found in caves, is essential for full understanding of their contained archaeology.