Body In Vertical Direction Research Articles

Omnidirectional Feature Learning for Person Re-Identification

Person re-identification (PReID) has received increasing attention due to it is an important part in intelligent surveillance. Recently, many state-of-the-art methods on PReID are part-based deep models. Most of them focus on learning the part feature representation of person body in horizontal direction. However, the feature representation of body in vertical direction is usually ignored. Besides, the spatial information between these part features and the different feature channels is not considered. In this study, we introduce a multi-branches deep model for PReID. Specifically, the model consists of five branches. Among the five branches, two of them learn the local feature with spatial information from horizontal or vertical orientations, respectively. The other one aims to learn interdependencies knowledge between different feature channels generated by the last convolution layer. The remains of two other branches are identification and triplet sub-networks, in which the discriminative global feature and a corresponding measurement can be learned simultaneously. All the five branches can improve the representation learning. We conduct extensive comparative experiments on three PReID benchmarks including CUHK03, Market-1501 and DukeMTMC-reID. The proposed deep framework outperforms many state-of-the-art in most cases.

21406 不整地路面走行時における人間-自転車系の力学(GS-28【最適化・計測(1)】)

This study deals with dynamic models of a human body and a bicycle during off-road cycling for optimization of the vibration characteristics of bicycle frame. A stability of a bicycle and a ride comfort depend on the characteristics of human-bicycle system. In addition, the mass ratio of a human body to a bicycle is extremely large as compared with other vehicles. Therefore, rider's posture becomes very important. In the present paper, the modal frequency and damping ratio of first mode of human body in vertical direction when the bicycle goes over an obstacle are estimated by experimental modal analysis. Especially, influence of condition of arms is clarified.

Biomechanics of human bipedal gallop: asymmetry dictates leg functions

Unilateral skipping or bipedal galloping is one of the gait types that humans are able to perform. In contrast to many animals, where gallop is the preferred gait at higher speeds, human bipedal gallop only occurs spontaneously in very specific conditions (e.g. fast downhill locomotion). This study examines the lower limb mechanics and explores the possible reasons why humans do not spontaneously opt for gallop for steady-state locomotion on level ground. In 12 subjects, who were required to run and gallop overground at their preferred speed, kinematic and kinetic data were collected and mechanical work at the main lower limb joints (hip, knee, ankle) was calculated. In a separate treadmill experiment, metabolic costs were measured. Analysis revealed that the principal differences between running and galloping are located at the hip. The asymmetrical configuration of gallop involves distinct hip actions and foot placing, giving galloping legs different functions compared with running legs: the trailing leg decelerates the body in the vertical direction but propels it forward while the leading leg acts in the opposite way. Although both legs conserve mechanical energy by interchanging external mechanical energy with potential elastic energy, the specific orientation of the legs causes more energy dissipation and generation compared with running. This makes gallop metabolically more expensive and involves high muscular stress at the hips, which may be why humans do not use gallop for steady-state locomotion.

Mechanical impedance of the human body in vertical direction

The mechanical impedance of the human body in sitting posture and vertical direction was measured during different experimental conditions, such as vibration level (0.5–1.4 m/s 2), frequency (2–100 Hz), body weight (57–92 kg), relaxed and erect upper body posture. The outcome shows that impedance increases with frequency up to a peak at about 5 Hz after which it decreases in a complex manner which includes two additional peaks. The frequency at which the first and second impedance peak occurs decreases with higher vibration level. Erect, compared with relaxed body posture resulted in higher impedance magnitudes and with peaks located at somewhat higher frequencies. Heavy persons show higher impedance magnitudes and peaks at lower frequencies.