• Annual Conference of KIPS
• /
• 2023.11a
• /
• pp.528-529
• /
• 2023

Due to vast points and irregular structure, labeling full points in large-scale point clouds is highly tedious and time-consuming. To resolve this issue, we propose a novel point-based transformer network in weakly-supervised semantic segmentation, which only needs 0.1% point annotations. Our network introduces general local features, representing global factors from different neighborhoods based on their order positions. Then, we share query point weights to local features through point attention to reinforce impacts, which are essential in determining sparse point labels. Geometric encoding is introduced to balance query point impact and remind point position during training. As a result, one point in specific local areas can obtain global features from corresponding ones in other neighborhoods and reinforce from its query points. Experimental results on benchmark large-scale point clouds demonstrate our proposed network's state-of-the-art performance.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.38B no.1
• /
• pp.1-9
• /
• 2013

This thesis proposes an automatic instruction extension system that utilizes retargetable compiler, based on MDL, to design an ASIP optimized for application. The proposed system uses information gathered from the application program to find all possible expandable instruction candidates. Expandable instruction candidates acquire the realization characteristics through hardware library. The system chooses instruction set and optimizes processor structure satisfying constraints on the bases of hardware characteristics and increase in execution speed. To confirm the efficiency of the proposed system, automatic instruction extension system was performed using various benchmark applications. The proposed system acquired optimized instruction set and processor structure, which are expanded from the commercial version of ARM9TDMI. Experimental results show that number of execution cycle has been reduced by 33.5% when compared to conventional version of ARM9TDMI, while area has been slightly increased.

• Journal of KIISE:Software and Applications
• /
• v.31 no.1
• /
• pp.61-70
• /
• 2004

It is difficult to find an appropriate web site because exponentially growing web contains millions of web documents. Personalization of web search can be realized by recommending proper web sites using user profile but more efficient method is needed for estimating preference because user's evaluation on web contents presents many aspects of his characteristics. As user profile has a property of non-linearity, estimation by classifier is needed and combination of classifiers is necessary to anticipate diverse properties. Structure adaptive self-organizing map (SASOM) that is suitable for Pattern classification and visualization is an enhanced model of SOM and might be useful for web mining. Fuzzy integral is a combination method using classifiers' relevance that is defined subjectively. In this paper, estimation of user profile is conducted by using ensemble of SASOM's teamed independently based on fuzzy integral and evaluated by Syskill ＆ Webert UCI benchmark data. Experimental results show that the proposed method performs better than previous naive Bayes classifier as well as voting of SASOM's.

• The Transactions of The Korean Institute of Electrical Engineers
• /
• v.57 no.1
• /
• pp.128-136
• /
• 2008

In this paper, we propose a new cache structure for effective error correction of soft error. We added check bit and SEEB(soft error evaluation block) to evaluate the status of cache line. The SEEB stores result of parity check into the two-bit shit register and set the check bit to '1' when parity check fails twice in the same cache line. In this case the line where parity check fails twice is treated as a vulnerable to soft error. When the data is filled into the cache, the new replacement algorithm is suggested that it can only use the valid block determined by SEEB. This structure prohibits the vulnerable line from being used and contributes to efficient use of cache by the reuse of line where parity check fails only once can be reused. We tried to minimize the side effect of the proposed cache and the experimental results, using SPEC2000 benchmark, showed 3% degradation in hit rate, 15% timing overhead because of parity logic and 2.7% area overhead. But it can be considered as trivial for SEEB because almost tolerant design inevitably adopt this parity method even if there are some overhead. And if only parity logic is used then it can have $5%{\sim}10%$ advantage than ECC logic. By using this proposed cache, the system will be protected from the threat of soft error in cache and the hit rate can be maintained to the level without soft error in the cache.

• Earthquakes and Structures
• /
• v.11 no.4
• /
• pp.583-607
• /
• 2016

One of the main applications of seismic risk assessment is that an specific design could be selected for a bridge from different alternatives by considering damage losses alongside primary construction costs. Therefore, in this paper, the focus is on selecting the shape of pylon, which is a changeable component in the design of a cable-stayed bridge, as a double criterion decision-making problem. Different shapes of pylons include H, A, Y, and diamond shape, and the two criterion are construction costs and probable earthquake losses. In this research, decision-making is performed by using developed seismic risk assessment process as a powerful method. Considering the existing uncertainties in seismic risk assessment process, the combined incremental dynamic analysis (IDA) and uniform design (UD) based fragility assessment method is proposed, in which the UD method is utilized to provide the logical capacity models of the structure, and the IDA method is employed to give the probabilistic seismic demand model of structure. Using the aforementioned models and by defining damage states, the fragility curves of the bridge system are obtained for the different pylon shapes usage. Finally, by combining the fragility curves with damage losses and implementing the proposed cost-loss-benefit (CLB) method, the seismic risk assessment process is developed with financial-comparative approach. Thus, the optimal shape of the pylon can be determined using double criterion decision-making. The final results of decision-making study indicate that the optimal pylon shapes for the studied span of cable-stayed bridge are, respectively, H shape, diamond shape, Y shape, and A shape.

• Journal of the Earthquake Engineering Society of Korea
• /
• v.9 no.4 s.44
• /
• pp.43-54
• /
• 2005

This paper presents cost-effectiveness evaluation of semi-active control system for cable-stayed bridge under earthquake excitations with various magnitudes and frequency contents. Semi-active control system, which is operated by using Bi-stale control method on the basis of linear quadratic Gaussian (LQG) optimal controller, is designed for the benchmark control problem proposed by Dyke et at. The cost-effectiveness of the proposed control system is defined by the ratio of life-cycle costs between a bridge structure with shock transmission units and a bridge structure with the semi-active control devices. The simulated results show that the damper cost has little influence on the cost-effectiveness of the semi-active control system while the cost-effectiveness is quite sensitive to the damage cost induced by the bridge failure. It is also found that the semi-active control system guarantees relatively high cost-effectiveness for the cable-stayed bridge subject to the ground motions in the regions of moderate seismicity with soft soil condition and strong seismicity with stiff soil condition.

• Journal of the Korean Institute of Telematics and Electronics C
• /
• v.36C no.11
• /
• pp.75-83
• /
• 1999

We propose a PfSGA(parameter-free species genetic algorithm) to learn the topological structure of MLP classifiers being adequate to given applications. The PfSGA is a combinational method of SGA(species genetic algorithm) and PfGA(parameter-free genetic algorithm). In SGA, we divide the total search space into several subspaces(species) according to the number of hidden units, and reduce the unnecessary search by eliminating the low promising species from the evolutionary process. However the performances of SGA classifiers are readily affected by the values of parameters such as mutation ratio and crossover ratio. In this paper, therefore, we combine SGA with PfGA, for which it is not necessary to determine the learning parameters. Experimental results on benchmark data and sign language words show that PfSGA can reduce the learning time of SGA and is not affected by the selection parameter values on structural learning. The results also show that PfSGA is more efficient than the exisiting methods in the aspect of misclassification ratio, learning rate, and complexity of MLP structure.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.12 no.1
• /
• pp.413-435
• /
• 2018

Although the accuracy of handwritten character recognition based on deep networks has been shown to be superior to that of the traditional method, the use of an overly deep network significantly increases time consumption during parameter training. For this reason, this paper took the training time and recognition accuracy into consideration and proposed a novel handwritten character recognition algorithm with newly designed network structure, which is based on an extended nonlinear kernel residual network. This network is a non-extremely deep network, and its main design is as follows:(1) Design of an unsupervised apriori algorithm for intra-class clustering, making the subsequent network training more pertinent; (2) presentation of an intermediate convolution model with a pre-processed width level of 2;(3) presentation of a composite residual structure that designs a multi-level quick link; and (4) addition of a Dropout layer after the parameter optimization. The algorithm shows superior results on MNIST and SVHN dataset, which are two character benchmark recognition datasets, and achieves better recognition accuracy and higher recognition efficiency than other deep structures with the same number of layers.

• Structural Engineering and Mechanics
• /
• v.63 no.2
• /
• pp.147-160
• /
• 2017

In this study, the optimum parameters of Tuned Mass Dampers (TMDs) are proposed using Gravity Search Algorithm (GSA) and Particle Swarm Optimization (PSO) to reduce the responses of the structures. A MATLAB program is developed to apply the new approach to the benchmark 10 and 40-story structures. The obtained results are compared to those of other optimization methods used in the literature to verify the developed code. To show the efficiency and accuracy of the proposed methods, nine far-field and near-field worldwide earthquakes are applied to the structures. The results reveal that in the 40-story structure, GSA algorithm can reduce the Relative Displacement (RD) and Absolute Acceleration (AA) up to 43% and 21%, respectively while the PSO decreases them by 50% and 25%, respectively. In contrast, both GSA and PSO algorithms reduce the RD and AA about 29% and 21% for the 10-story structure. Furthermore, using the proposed approach the required TMD parameters reduce by 47% and 63% in the 40 and 10-story buildings in comparison with the referenced ones. Result evaluation and related comparison indicate that these methods are more effective even by using smaller TMD parameters resulting in the reduction of acting force from TMD, having smaller stiffness and damping factors while being more cost effective due to its decreased parameters. In other words, the TMD with optimum parameters can play a positive role in both tall and typical structures.

• Smart Structures and Systems
• /
• v.21 no.3
• /
• pp.359-371
• /
• 2018

Strong seismic events commonly cause large drift and deformation, and functionality failures in the superstructures. One way to prevent functionality failures is to design structures which are ductile and flexible through yielding when subjected to strong ground excitations. By developing forces that assist motion as "negative stiffness forces", yielding can be achieved. In this paper, we adopt the weakening and damping method to achieve a new approach to reduce all of the structural responses by further adjusting damping phase. A semi-active control system is adopted to perform the experiments. In this adaptation, negative stiffness forces through certain devices are used in weakening phase to reduce structural strength. Magneto-rheological (MR) dampers are then added to preserve stability of the structure. To adjust the voltage in MR dampers, an inverse model is employed in the control system to command MR dampers and generate the desired control forces, where a velocity control algorithm produces initial required control force. An extensive numerical study is conducted to evaluate proposed methodology by using the smart base-isolated benchmark building. Totally, nine control systems are examined to study proposed strategy. Based on the numerical results of seven earthquakes, the use of proposed strategy not only reduces base displacements, base accelerations and base shear but also leads to reduction of accelerations and inter story drifts of the superstructure. Numerical results shows that the usage of inverse model produces the desired regulated damping, thus improving the stability of the structure.