Image recognition can be applied in many applications of Intelligent Transportation System. Through automated traffic flow counting, the traffic information can be presented effectively for a given area. After the existing image recognition model process the monitoring video, the coordinates of objects in each frame can be easily extracted. The extracted object coordinates are then filtered to obtain the required vehicle coordinates. To achieve the function of vehicle counting, it is necessary to identify the relationship of vehicles in different frames, i.e., whether or not they represent the same vehicle. Although the vehicle counting can be achieved by using the tracking algorithm, a short period of recognition failure may cause wrong tracking, which will lead to incorrect traffic counting. In this paper, we propose a system that utilizes the YOLO framework for traffic flow counting. The system architecture consists of three blocks, including the Detector that generates the bounding box of vehicles, the Buffer which stores coordinates of vehicles, and the Counter which is responsible for vehicle counting. The proposed system requires only to utilize simple distance calculations to achieve the purpose of vehicle counting. In addition, by adding checkpoints, the system is able to alleviate the consequence of false detection. The videos from different locations and angles are used to verify and analyze the correctness and overall efficiency of the proposed system, and the results indicate that our system achieves high counting accuracy under the environment with sufficient ambient light.
{"title":"A YOLO-Based Traffic Counting System","authors":"Jia-Ping Lin, Min-Te Sun","doi":"10.1109/TAAI.2018.00027","DOIUrl":"https://doi.org/10.1109/TAAI.2018.00027","url":null,"abstract":"Image recognition can be applied in many applications of Intelligent Transportation System. Through automated traffic flow counting, the traffic information can be presented effectively for a given area. After the existing image recognition model process the monitoring video, the coordinates of objects in each frame can be easily extracted. The extracted object coordinates are then filtered to obtain the required vehicle coordinates. To achieve the function of vehicle counting, it is necessary to identify the relationship of vehicles in different frames, i.e., whether or not they represent the same vehicle. Although the vehicle counting can be achieved by using the tracking algorithm, a short period of recognition failure may cause wrong tracking, which will lead to incorrect traffic counting. In this paper, we propose a system that utilizes the YOLO framework for traffic flow counting. The system architecture consists of three blocks, including the Detector that generates the bounding box of vehicles, the Buffer which stores coordinates of vehicles, and the Counter which is responsible for vehicle counting. The proposed system requires only to utilize simple distance calculations to achieve the purpose of vehicle counting. In addition, by adding checkpoints, the system is able to alleviate the consequence of false detection. The videos from different locations and angles are used to verify and analyze the correctness and overall efficiency of the proposed system, and the results indicate that our system achieves high counting accuracy under the environment with sufficient ambient light.","PeriodicalId":211734,"journal":{"name":"2018 Conference on Technologies and Applications of Artificial Intelligence (TAAI)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2018-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121616031","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
In this paper, we propose a weighted voting method for a heterogeneous game system, which assigns the strength of engines and win probabilities of their positions to the weights for voting. Assigning the strength as the weight solves the problem of weaker engines entering the majority voting. The win probabilities are transformed from the evaluation values by a sigmoid function generated for each engine. Through the sigmoid functions, we can compare the win probabilities between the different engines and resolve the problem of optimistic voting in heterogeneous systems. Optimistic voting, which simply selects the highest-scoring move, may select a suboptimal random move when random players are involved in the game. Finally, we competed the proposed system and other voting systems against a single engine in shogi tournaments and compared the strengths of the systems in shogi. The experimental results confirmed the effectiveness of the proposed method.
{"title":"Weighted Majority Voting with a Heterogeneous System in the Game of Shogi","authors":"Shogo Takeuchi","doi":"10.1109/TAAI.2018.00035","DOIUrl":"https://doi.org/10.1109/TAAI.2018.00035","url":null,"abstract":"In this paper, we propose a weighted voting method for a heterogeneous game system, which assigns the strength of engines and win probabilities of their positions to the weights for voting. Assigning the strength as the weight solves the problem of weaker engines entering the majority voting. The win probabilities are transformed from the evaluation values by a sigmoid function generated for each engine. Through the sigmoid functions, we can compare the win probabilities between the different engines and resolve the problem of optimistic voting in heterogeneous systems. Optimistic voting, which simply selects the highest-scoring move, may select a suboptimal random move when random players are involved in the game. Finally, we competed the proposed system and other voting systems against a single engine in shogi tournaments and compared the strengths of the systems in shogi. The experimental results confirmed the effectiveness of the proposed method.","PeriodicalId":211734,"journal":{"name":"2018 Conference on Technologies and Applications of Artificial Intelligence (TAAI)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2018-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116966057","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Game 2048 is a stochastic single-player game and development of strong computer players for 2048 has been based on N-tuple networks trained by reinforcement learning. In our previous study, we developed computer players for game 2048 based on convolutional neural networks (CNNs), and showed by experiments that networks with three or more convolution layers performed much better than that with two convolution layers. In this study, we analyze the inner working of our CNNs (i.e. white box approach) to identify the reasons of the performance. Our analyses include visualization of filters in the first layers and backward trace of the networks for some specific game states. We report several findings about inner working of our CNNs for game 2048.
Game 2048是一个随机的单人游戏,为2048开发强大的计算机玩家是基于强化学习训练的n元网络。在我们之前的研究中,我们基于卷积神经网络(cnn)开发了游戏2048的计算机播放器,并通过实验表明,具有三个或更多卷积层的网络比具有两个卷积层的网络性能要好得多。在本研究中,我们分析了我们的cnn的内部工作(即白盒方法),以确定性能的原因。我们的分析包括对第一层过滤器的可视化和对某些特定游戏状态的网络的反向跟踪。我们报告了关于2048场比赛cnn内部工作的一些发现。
{"title":"Interpreting Neural-Network Players for Game 2048","authors":"Kiminori Matsuzaki, Madoka Teramura","doi":"10.1109/TAAI.2018.00038","DOIUrl":"https://doi.org/10.1109/TAAI.2018.00038","url":null,"abstract":"Game 2048 is a stochastic single-player game and development of strong computer players for 2048 has been based on N-tuple networks trained by reinforcement learning. In our previous study, we developed computer players for game 2048 based on convolutional neural networks (CNNs), and showed by experiments that networks with three or more convolution layers performed much better than that with two convolution layers. In this study, we analyze the inner working of our CNNs (i.e. white box approach) to identify the reasons of the performance. Our analyses include visualization of filters in the first layers and backward trace of the networks for some specific game states. We report several findings about inner working of our CNNs for game 2048.","PeriodicalId":211734,"journal":{"name":"2018 Conference on Technologies and Applications of Artificial Intelligence (TAAI)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2018-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124582868","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The TAAI conference is an annual event that brings together researchers, engineers, and practitioners to present and exchange ideas, results, and experience in AI technologies and applications. This year, the conference received paper submissions from different countries. Every submission was assigned to at least two experts from the technical program committee to review. Due to the large amount of high-quality submissions, regular paper acceptance was very competitive. In addition, these proceedings feature high-quality papers. All of these papers provide novel ideas, new results, and state-of-the-art techniques in the field. We are honored to have several of the world’s leading experts in the field join us as distinguished keynote speakers. Altogether, we are proud to be able to present you a rich program that contains a variety of excellent researches.
{"title":"Message from the TAAI 2018 Program Co-Chairs","authors":"Chao-Tung Yang, Chao Chen","doi":"10.1109/taai.2018.00006","DOIUrl":"https://doi.org/10.1109/taai.2018.00006","url":null,"abstract":"The TAAI conference is an annual event that brings together researchers, engineers, and practitioners to present and exchange ideas, results, and experience in AI technologies and applications. This year, the conference received paper submissions from different countries. Every submission was assigned to at least two experts from the technical program committee to review. Due to the large amount of high-quality submissions, regular paper acceptance was very competitive. In addition, these proceedings feature high-quality papers. All of these papers provide novel ideas, new results, and state-of-the-art techniques in the field. We are honored to have several of the world’s leading experts in the field join us as distinguished keynote speakers. Altogether, we are proud to be able to present you a rich program that contains a variety of excellent researches.","PeriodicalId":211734,"journal":{"name":"2018 Conference on Technologies and Applications of Artificial Intelligence (TAAI)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2018-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131779009","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Tongue diagnosis is a unique practice in traditional Chinese medicine(TCM), which can be used to infer the health condition of a person. However, different TCM doctors may give different interpretations on the same tongue. If an artificial intelligence model can be developed based on a large number of doctor-interpreted tongue images, a more objective judgment will be obtained. Deep learning in artificial intelligence has excellent performance in image recognition, and feature extraction can be done automatically by deep learning without image processing experts. This study attempts to develop a deep learning model through a large number of tongue images, especially for tongue fissures. We also visualize the fissure regions with Gradient-weighted Class Activation Mapping(Grad-cam). Therefore, the model not only try to detect tongue fissures but also localize tongue fissure regions.
{"title":"Tongue Fissure Visualization with Deep Learning","authors":"Wen-Hsien Chang, H. Chu, Hen-Hong Chang","doi":"10.1109/TAAI.2018.00013","DOIUrl":"https://doi.org/10.1109/TAAI.2018.00013","url":null,"abstract":"Tongue diagnosis is a unique practice in traditional Chinese medicine(TCM), which can be used to infer the health condition of a person. However, different TCM doctors may give different interpretations on the same tongue. If an artificial intelligence model can be developed based on a large number of doctor-interpreted tongue images, a more objective judgment will be obtained. Deep learning in artificial intelligence has excellent performance in image recognition, and feature extraction can be done automatically by deep learning without image processing experts. This study attempts to develop a deep learning model through a large number of tongue images, especially for tongue fissures. We also visualize the fissure regions with Gradient-weighted Class Activation Mapping(Grad-cam). Therefore, the model not only try to detect tongue fissures but also localize tongue fissure regions.","PeriodicalId":211734,"journal":{"name":"2018 Conference on Technologies and Applications of Artificial Intelligence (TAAI)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2018-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131644344","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"[Copyright notice]","authors":"","doi":"10.1109/taai.2018.00003","DOIUrl":"https://doi.org/10.1109/taai.2018.00003","url":null,"abstract":"","PeriodicalId":211734,"journal":{"name":"2018 Conference on Technologies and Applications of Artificial Intelligence (TAAI)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2018-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124307492","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Werewolf, also known as Mafia, is a kind of game with imperfect information. Werewolf game agents must cope with two kinds of problems, "decision on who to trust or to kill", and "decision on information exchange". In this paper, we focus on the first problem. We apply techniques in Deep Q Network in building werewolf agents. We also improve representation of states and actions based on existing agents trained by Q learning method. Our proposed agents were compared with existing agents trained by Q learning method and with existing agents submitted to the AIWolf Contest, the most famous werewolf game agents contest in Japan. For every role, we prepared four agents with proposed method and investigated average win ratio of four agents in our experiments. Experimental results showed that when agents learned and played with the same group of players, our proposed agents have better player performances than existing agents trained by Q learning method and a part of agents submitted to the AIWolf Contest. We obtained promising results by using reinforcement learning method to solve "decision on who to trust or to kill" problem without using heuristic methods.
{"title":"Application of Deep Reinforcement Learning in Werewolf Game Agents","authors":"Tianhe Wang, Tomoyuki Kaneko","doi":"10.1109/TAAI.2018.00016","DOIUrl":"https://doi.org/10.1109/TAAI.2018.00016","url":null,"abstract":"Werewolf, also known as Mafia, is a kind of game with imperfect information. Werewolf game agents must cope with two kinds of problems, \"decision on who to trust or to kill\", and \"decision on information exchange\". In this paper, we focus on the first problem. We apply techniques in Deep Q Network in building werewolf agents. We also improve representation of states and actions based on existing agents trained by Q learning method. Our proposed agents were compared with existing agents trained by Q learning method and with existing agents submitted to the AIWolf Contest, the most famous werewolf game agents contest in Japan. For every role, we prepared four agents with proposed method and investigated average win ratio of four agents in our experiments. Experimental results showed that when agents learned and played with the same group of players, our proposed agents have better player performances than existing agents trained by Q learning method and a part of agents submitted to the AIWolf Contest. We obtained promising results by using reinforcement learning method to solve \"decision on who to trust or to kill\" problem without using heuristic methods.","PeriodicalId":211734,"journal":{"name":"2018 Conference on Technologies and Applications of Artificial Intelligence (TAAI)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2018-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125448800","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Message from the TAAI 2018 General Co-Chairs","authors":"","doi":"10.1109/taai.2018.00005","DOIUrl":"https://doi.org/10.1109/taai.2018.00005","url":null,"abstract":"","PeriodicalId":211734,"journal":{"name":"2018 Conference on Technologies and Applications of Artificial Intelligence (TAAI)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2018-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116782467","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
As shown in AlphaGo, AlphaGo Zero, and AlphaZero, reinforcement learning is effective in learning of evaluation functions (or value networks) in Go, Chess and Shogi. In their training, two procedures are repeated in parallel; self-play with a current evaluation function and improvement of the evaluation function by using game records yielded by recent self-play. Although AlphaGo, AlphaGo Zero, and AlphaZero have achieved super human performance, the method requires enormous computation resources. To alleviate the problem, this paper proposes to incorporate a checkmate solver in self-play. We show that this small enhancement dramatically improves the efficiency of our experiments in Minishogi, via the quality of game records in self-play. It should be noted that our method is still free from human knowledge about a target domain, though the implementation of checkmate solvers is domain dependent.
{"title":"Learning of Evaluation Functions via Self-Play Enhanced by Checkmate Search","authors":"T. Nakayashiki, Tomoyuki Kaneko","doi":"10.1109/TAAI.2018.00036","DOIUrl":"https://doi.org/10.1109/TAAI.2018.00036","url":null,"abstract":"As shown in AlphaGo, AlphaGo Zero, and AlphaZero, reinforcement learning is effective in learning of evaluation functions (or value networks) in Go, Chess and Shogi. In their training, two procedures are repeated in parallel; self-play with a current evaluation function and improvement of the evaluation function by using game records yielded by recent self-play. Although AlphaGo, AlphaGo Zero, and AlphaZero have achieved super human performance, the method requires enormous computation resources. To alleviate the problem, this paper proposes to incorporate a checkmate solver in self-play. We show that this small enhancement dramatically improves the efficiency of our experiments in Minishogi, via the quality of game records in self-play. It should be noted that our method is still free from human knowledge about a target domain, though the implementation of checkmate solvers is domain dependent.","PeriodicalId":211734,"journal":{"name":"2018 Conference on Technologies and Applications of Artificial Intelligence (TAAI)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2018-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123165547","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Yang Liu, Guijuan Zhang, Xiaoning Jin, Yaozong Jia
The personalized video recommendation system provides users with great convenience while surfing in the video websites. Among many algorithms adopted by recommendation system, the collaborative filtering algorithm is the most widely used and has achieved great success in practical applications, however, the recommended performance suffers from the problem of data sparsity severely. We propose a model that adopts Doc2Vec to deal with video's text information and integrates genre information into rating matrix pre-padding to reduce the sparsity of ratings. The experimental results show that pre-padding ratings is of high quality and the algorithms based on collaborative filtering achieve better performance on the padded datasets.
{"title":"Rating Matrix Pre-Padding for Video Recommendation","authors":"Yang Liu, Guijuan Zhang, Xiaoning Jin, Yaozong Jia","doi":"10.1109/TAAI.2018.00044","DOIUrl":"https://doi.org/10.1109/TAAI.2018.00044","url":null,"abstract":"The personalized video recommendation system provides users with great convenience while surfing in the video websites. Among many algorithms adopted by recommendation system, the collaborative filtering algorithm is the most widely used and has achieved great success in practical applications, however, the recommended performance suffers from the problem of data sparsity severely. We propose a model that adopts Doc2Vec to deal with video's text information and integrates genre information into rating matrix pre-padding to reduce the sparsity of ratings. The experimental results show that pre-padding ratings is of high quality and the algorithms based on collaborative filtering achieve better performance on the padded datasets.","PeriodicalId":211734,"journal":{"name":"2018 Conference on Technologies and Applications of Artificial Intelligence (TAAI)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2018-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116125211","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}