This paper proposes a method to count vehicles from an in-vehicle camera image by regression based on car parts detection. In the case of an in-vehicle camera image, since vehicles are frequently occluded by other vehicles in traffic congestion, it is difficult to accurately count vehicles. Therefore, we propose a method to count vehicles by regression based on the number of visible car parts. For this, we make an estimator by learning the relation between the number of visible car parts and that of vehicles by Support Vector Regression. We evaluated our method using in-vehicle camera images recorded in an actual environment, where the proposed method performed better than counting detected vehicles.

In real-world driving, one of the critical issues to be analyzed is an interaction with other traffic participants. This problem is easily solved in the case that direct communication between agents is available. Such direct communication, however, is not always available due to limited implementation of realtime on-board communication facilities. For sophisticated safety system design, the vehicle is highly requested to implement an interactive intelligence, which is mainly based on an understanding of interaction mechanism between traffic participants from observed sensing signals. In order to analyze the interactive behavior, this paper presents a modeling by using a PieceWise AutoRegressive eXogenous (PWARX) model. Since the PWARX model can describe continuous dynamics and discrete switching, interactive behavior is modeled and understood as the combination of primitive dynamics (operation) and mode transitions (decision making) of them. The usefulness of the proposed scheme is verified by applying to the modeling of interactive task of ‘bidirectional passing by task.’

Recently, practical applications of automatic driving have been rapidly developing. In the future, it must be necessary to enable interactive operation by natural language in order to easily operate autonomous cars. We therefore attempt to realize the correspondence relationship, i.e., a part of symbol grounding, between the driving instructions expressed in natural language and the objects in the real world recognized by the sensors equipped with a car, and then convert the driving instructions into the particular spatial meaning description to operate autonomous cars. In this study, we particularly focus on the parking operation of a car. We propose two methods: one is extracting spatial semantics from parking instructions, and the other is corresponding spatial semantics with the real-world environment. The structure trees given by Combinatory Categorial Grammar (CCG) are used as intermediate representation in exacting spatial semantics. If unknown words appear, we estimate them by using Conditional Random Field. In order to increase the accuracy of CCG parser, we implement a reranker of parse trees. These parse trees are converted into tree structures called Spatial Description Clause (SDC). We extend the framework of SDC by adding two new semantic categories, VIEW and STATE, so as to be able to ground more variety of the instructions for driving a car in the real-world environment. In corresponding spatial semantics with the real-world environment, we generate probability graphical models called Generalized Grounding Graph and output places or objects which correspond each word. The accuracy of all grounding among the sentences correctly parsed is 79.2%.

Non-invasive brain-computer interface (BCI) based on electroencephalogram (EEG) is generally limited to specific measurement sites and frequency bands. These types of BCI systems utilize certain target EEG features evoked by cognitive tasks and detects the certain EEG features to determine BCI control. However, it seems that the user who well reproduce these EEG features are suitable for such EEG-BCI, but often the BCI is not suitable for other users without reproducible EEG features evoked by the same task. Then we propose a heuristic BCI system without a priori assumption of focused EEG features for detection. In this system, measurement sites and frequency bands are not selected in advance, but the BCI system searches the suitable measurement sites and frequency bands suitable for reproducible EEG features for the user, by learning process. Learning-type Fuzzy Template Matching method (L-FTM) based on simplified fuzzy logic with learning is used for the heuristic algorithm. We succeeded in extracting the EEG features evoked during right-upper limb motor imagery task, using the developed BCI system. We also confirmed that using the heuristic-BCI system, with fuzzy rules corresponding to specific EEG feature patterns were automatically extracted by the learning process. Thus, the developed BCI system was suggested that it learns and extracts EEG features corresponding to a certain task without prior information. The BCI system can also be used as an effective tool for detecting EEG features.