• Journal of Broadcast Engineering
• /
• v.27 no.3
• /
• pp.318-331
• /
• 2022

With the recent development of hardware computing devices and software based frameworks, machine tasks using deep learning networks are expected to be utilized in various industrial fields and personal IoT devices. However, in order to overcome the limitations of high cost device for utilizing the deep learning network and that the user may not receive the results requested when only the machine task results are transmitted from the server, Collaborative Intelligence (CI) proposed the transmission of feature maps as a solution. In this paper, an efficient compression method for feature maps with vast data sizes to support the CI paradigm was analyzed and presented through experiments. This method increases redundancy by applying feature map reordering to improve compression efficiency in traditional video codecs, and proposes a feature map method that improves compression efficiency and maintains the performance of machine tasks by simultaneously utilizing image compression format and video compression format. As a result of the experiment, the proposed method shows 14.29% gain in BD-rate of BPP and mAP compared to the feature compression anchor of MPEG-VCM.

• The korean journal of orthodontics
• /
• v.28 no.5 s.70
• /
• pp.813-823
• /
• 1998

The frictional force has been considered as an harmful factor in an active unit where tooth movement occurs, but as an advantageous factor in anchor unit that resist tooth movement. That is, efficient tooth movement is planned by using ligation methods that have low levels of bracket-wire frictional force and the anchorage control can be achieved by using ligation methods that have high levels of bracket-wire frictional force that result in binding of the bracket accompanied by little or no tooth movement. The purpose of this study was to evaluate the frictional force generated between bracket and wire in accordance with the methods of ligation, the material of ligation and the passage of time under artificial saliva. Tested were 0.017x0.022 inch stainless steel wires in standard edgewise twin brackets for upper central incisors in a 0.018-inch slot. The wires were ligated into the brackets with elastomeric modules and stainless steel ligatures. Whole tie, half tie, twisting tie and double overlay tie were done with elastomeric modules. With 0.009-inch stainless steel ligature whole tie and half tie were done by needle holder and whole tie by ligature tying plier. With 0.012-inch stainless steel ligature whole ties were done by needle holder. Whole tie groups of elastomeric module were kept in artificial saliva bath at $37^{\circ}C$ for 28 days. The frictional force was recorded by means of an Instron universial testing instrument (4202 INSTRON, Instron Co., U.S.A.) at initial, 7, 14, 21, and 28 days. The results for ligated samples in a simulated oral environment revealed the fellowing : ${\cdot}$In elastomeric module whole tie, 28 days group was significantly greater mean static frictional force than any other group but there were no significant differences among any other group (p>0.05). ${\cdot}$Elastomeric module twisting ties were significantly greater mean static frictional forces than any other ligation method but there were no significant differences between twisting tie and double overlay tie (p>0.05). Twisting tie, double overlay tie, whole tie, half tie showed differences in decreasing order. ${\cdot}$Stainless steel half tie produced lower mean static frictional force than whole tie, ligation by ligature tying plier produced greater mean static frictional force than by needle holder and ligation with 0.012-inch stainless steel ligature produced greater mean static frictional force than with 0.009-inch stainless steel ligature (p<0.05). ${\cdot}$There were no significant differences between the mean static frictional forces of elastomeric whole tie and stainless steel whole tie (p>0.05).