Earticle

This study is aimed at proposing the following two possibilities and analyzing them in connection with the positions of objects in Old French(OF) (11c-12c): i) Can objects move to the left position of a clause by means of Object Shift? and ii) Does the object raising comply with the feature inheritance theory proposed by Chomsky(2008) in order to realize the V2-structure of OF constructions? Although these two syntactic operations have quite close relations, we will verify which parameter can apply to them in OF considering that they operate in a different manner among languages. Consequently, we will observe and analyze the following subjects: i) The trigger in C is a tense feature which causes a finite verb to raise to C; ii) What moves an object to Spec-CP is an edge feature in C?; iii) How is the OF object raising different from the object shift?; iv) How can the status of lexical or null subject be determined during the OF object raising/shift?; v) Is the object raising/shift derived from the verb movement and the feature inheritance in OF? Shedding light on these tasks, we will clarify that the object raising/shift can be verified by the verb movement and feature inheritance in the object raising constructions observed in La Chanson de Roland(1090) and Le Chevalier de la Charrete(1179-81).

In GB framework, covert operations, such as covert raising of wh-in-situ and quantifier raising, had been postulated to explain such phenomena as ACD constructions or scope interactions between the scope bearing elements. However, the recent minimalist approaches may have no way of explaining them in the framework since there is no covert operations in narrow syntax. To solve the problems, I will propose such covert operations be reconsidered and that the wh-in-situ in English be raised overtly to the outer Spec of v*P to satisfy the EPP of v*P and further be raised beyond the position through DISL. By assuming this proposal, we can get several advantages to explain the nature of human languages. First, we reduce the computational burden raised in Chomsky's(2000, 2001) approach to wh-movement constructions. The object shift of wh-element to the outer Spec of v*P causes computational complexity in Chomsky's system since the raising of wh-element is triggered just by the observance of PIC. However, we have no difficulty raising wh-element to the outer Spec of v*P since wh-element must raise to satisfy the EPP of v*P. Second, we can capture the correct reading of multiple-wh-questions eliminating the asymmetry between overt wh-movement and wh-in-situ. Finally, we can derive the object shift property of wh-movement by assuming that the shifted element in the outer Spec of v*P further raises beyond the position through wh-movement or TH/EX.

Structural Case (i.e., inherent Case) is associated with theta-positions, whereas structural Case is not (Chomsky 1981, 1986, Bošković 1997, 2002, Woolford 2006). If inherent Case (which Bošković assumes is assigned to elements in situ) is available only to nominals that the Case-assigning verb θ-marks, as is standardly assumed, inherent Case is excluded from ECM constructions. The inherent Case option of the ECMed nominal is ruled out due to the association with θ-licensing, the ECMed nominal not being θ-marked by its Case licensor: the only Case available to objects that undergo ECM, then is structural. In this paper, following Bošković (1997, 2002), I argue that structural Case must be checked by movement. In this respect, the ECMed nominal must move to the matrix non-thematic position (i.e., here, the outer spec of the matrix v*) in order to have its structural Case checked against v*: that is, the Korean ECMed nominal receives its structural Case in its surface position. From that surface A-position, the ECMed nominal can be passivized to the matrix [Spec, TP]. Unlike in English, the embedded clause of ECM constructions in Korean has an overt tense morpheme and an overt complementizer, which indicates that it is of a CP structure. If we assume that the ECMed nominal moves from the embedded subject position, it must raise across the embedded [Spec, CP] by A-movement. This violates locality on A-movement.

In this paper, we propose a novel method for automatic object segmentation using Mean shift with Grid-mask for Grab Cut algorithm. The main idea of proposed method is a rapid and exact technique for extracting initial foreground information using Mean shift with Grid-mask then we make initial rectangle around estimated initial foreground. And then Grab Cut algorithm is applied to segment foreground form background based on initial rectangle provided by previous process. In order to evaluate our proposed method, we compare the proposed method with several competitive automatic methods on the MSRA database which has 1000 images with ground truth. The scheme successfully segments objects without prior knowledge with both higher precision and better recall than competitive methods.

An online mean-shift object tracking algorithm, which consists of a learning stage and an estimation stage, is proposed in this work. The learning stage selects the features for tracking, and the estimation stage composes a likelihood image and applies the mean shift algorithm to it to track an object. The tracking performance depends on the quality of the likelihood image. We propose two schemes to generate and integrate likelihood images: one based on the discrete AdaBoost (DAB) and the other based on the real AdaBoost (RAB). The DAB scheme uses tuned feature values, whereas RAB estimates class probabilities, to select the features and generate the likelihood images. Experiment results show that the proposed algorithm provides more accurate and reliable tracking results than the conventional mean shift tracking algorithms.

This article presents a novel analysis showing that Korean objects always move out of the VP where they are generated and often travel through hallmark positions of VP-adverbs. Although this analysis, referred to as short object shift in this article, appears to run counter to several constraints of movement, it will be shown to admit of a principled account if we make full use of the minimalist ideas behind Chomsky's (2013, 2015) labeling algorithm, in combination with the noun formation concepts of Borer (2005a, b), Sportiche (2005), and others. The combined system reduces clausal structures by dispensing with AgrOP, and more importantly derives the dual nature of quantifier float that has been controversial over the past decades. This path of the minimalist approach receives further credit in comparison with Boskovic’s (2004) analysis on the so-called ‘the Ban on Quantifier Float in Θ?positions(BQFT).’ Because, in the present system, full nouns are only available in derived positions but not in Θ?positions, it can naturally be expected that floating quantifiers, which require full nouns for interpretations, appear in non-Θ?positions.

This paper argues that the surface position of the ECMed nominal in Korean should be [Spec, v*P] in the upper clause. This paper is against the arguments that the Korean ECM construction is a syntactic reflection of pragmatic force (i.e., Int effects in Chomsky's (2001) term) (Jung 2001, Yoon 2004, among others). In fact, the overt object shift of the ECMed nominal is driven by its structural Case checking against the matrix v* (cf. Bo?kovi? 1997, 2002, in press a, Radford 2004). This paper also claims that while the object with inherent Case in the thematic position undergoes Case checking against v* by Agree (Chomsky 2000, 2001), the ECMed nominal with structural Case necessarily undergoes overt object shift in order to have its structural Case checked against v*. No instance of feature checking movement can feed another instance of feature checking movement (Bo?kovi? in press a, b). It makes sense that the ECM construction in Korean involves A-movement out of CP, which induces the [Spec, CP] to be an A-position (cf. Tanaka 2002, McCloskey 2000, Bo?kovi? in press a, b). It is also observed that the Korean ECMed nominal may further undergo a separated feature checking operation from its surface shifted position to another surface position; hence, no feature checking in intermediate positions in successive cyclic movement of the ECMed nominals.

Aoun and Li (1989) and Hornstein (1995) suggest that the cross-linguistic contrast in quantifier scope between English and East Asian languages is attributed to the parametric difference in the base subject position, viz. VP-internal position in English and Spec IP in East Asian languages. This paper argues that their suggestion is untenable, and that the cross linguistic contrast in question is due to the parametric difference that English permits and East Asian languages do not permit (overt) object shift.

This paper develops an argument for syntactic treatments on Object Shift in Scandinavian languages. In this paper object shift is considered to be a syntactic operation based on consideration of Holmberg's Generalization[HG], well -documented constraint on object shift in the Scandinavian language. Assumed Matushansky(2003)'s unification movement theory, I try to explain the properties of object shift, in particular aspects having to do with head-to-XP movement and [±Foc]features. The Unification Movement theory runs like this: Head-movement and phrasal movement assume to be same thing and head undergoes movement to SPEC of XP in a same way as XP does. From the SPEC of XP, it is argued that head-merging operation called M-Merger happens. In this paper, if object moves to the SPEC of XP, The objects also undergo head-merging operation with verbs under the rule specified in this paper, which formulates that [-Foc] units such as the weak pronouns in Scandinavian language should be in adjacency with [+Foc] units.

Since Holmberg (1986) first presented an analysis of object shift (OS) phenomenon observed in Icelandic, OS has been one of the topics that has intrigued many linguists including N. Chomsky. The purpose of the present paper is not to present a new analysis of OS but to show how OS is analyzed in the minimalist program, based on Chomsky(2001). I believe there are two points that have to be made clear in the analysis but not yet accounted for in any clear way: one is where the "shifted" object finally moves, and the other what is regarded as properties of the phonological component (i.e., phonological border) is used for the analysis of OS.

Quantifier floating (Q-floating) displays interesting asymmetries in English. First of all, there is a subject/object asymmetry. The subject permits Q-floating, whereas the object does not. However, if the object is followed by a predicative constituent, Q-floating can be permitted. In this case, there is another subject/object asymmetry. If the object is followed by a constituent that bears a predication relation with it, Q-floating is permitted, If, on the other hand, the object is accompanied by a constituent that bears a predication relation with the subject, Q-floating is not permitted. This paper shows that the various types of asymmetries follow if (i) Q-floating is licensed when A-movement takes place (Sportiche 1988), (ii) object can move to SPEC-V (Chomsky (2008, 2013, 2015), but in simple transitive constructions raising of the object to SPEC-V is prohibited by an anti-locality condition, and (iii) the movement theory of control is correct (Hornstein 1999, 2001).

Harley (2002)는 영어의 수여동사 give와 send는 술어 CAUSE와 HAVE/ LOCATION으로 각각 분해 (decompose)될 수 있고, 이러한 어휘적 분해는 표면적으로 이중목적어구문인 John gave Mary a book과 여격구문인 John gave a book to Mary라는 형태로 각각 표출된다고 주장한다. 한국어의 수여동사구문은 한국어가 뒤섞기 언어라는 특수성 때문에 영어와 유사한 두 가지 형태의 표면구조를 가지는가의 여부를 규명하기는 쉽지 않다. 이 문제에 대하여 Jung and Miyagawa (2004)는 한국어에도 영어와 유사한 두 가지 형태의 구문이 존재하며, 그 구문들의 특성도 기본적으로 영어와 동일함을 보였다. 그러나 Jung and Miyagawa (2004)가 관찰한 두 언어 간의 유사성에도 불구하고, 영어와 한국어는 이중목적어구문에 나타나는 간접목적어의 의미적 해석에 있어 중요한 차이점을 보인다. 즉, 한국어의 대격표지 간접목적어는 특정성이라는 의미해석을 의무적으로 가지는데 비해, 영어의 간접목적어는 그러한 의미제약이 없다. 본 논문은 영어와는 달리 한국어에서는 왜 그러한 특정성 제약이 있는가의 문제를 중점적으로 논의하였다. 결론적으로 본 논문은 한국어의 대격표지 간접목적어가 가지는 특정성 해석은 영어와는 달리 술어 vCAUSE가 이중 대격을 인허하는 과정에서 생기는 자연적 귀결임을 보여주었다.

Dae-Ik Kim. 2004. Verb Movement and Object Shift in Scandinavian. Studies in Modern Grammar 35, 1-20. This paper develops an argument for the M-merger of Movement based on consideration of Holmberg's Generalization[HG], well-documented constraint on object shift in the Scandinavian language. Assumed head-movement and phrasal movement are the same thing, Head undergoes movement to SPEC of XP in a same way as XP does. From the SPEC of XP, it is argued that head-merging operation called M-Merger happens. If Object is placed in the SPEC of XP, objects also undergo head-merging operation with verbs under rich morphological formations such as Icelandic full DP or Weak pronouns in mainland scandinavian. This operation assumed to be a morphological process showing a different pattern to that of syntax, which gives the reason why object shift shows the mixed nature in syntactic phenomena. It is argued also that the characterization of HG presented here can be explained in terms of M-Merger(Young San University)

The mean shift algorithm has achieved considerable success in object tracking due to its simplicity and robustness. It finds local maxima of a similarity measure between the color histograms of the target and candidate image. However, the mean shift tracking algorithm using only color histograms has a serious defect. It doesn't use the spatial information of the target. Thus, it is difficult to model the target more exactly. And it is likely to lose the target during the occlusions of other objects which have similar color distributions. To deal with these difficulties we use both color information and spatial information of the target. Our proposed algorithm is robust to occlusions and scale changes in front of dynamic, unstructured background. In addition, our proposed method is computationally efficient. Therefore, it can be executed in real-time.

이 논문에서는 CAM shift알고리즘과 8방향 탐색 위도우를 결합하여 객체의 추적 성능을 향상하는 방법과 추적에 이용되는 프레임의 수를 줄여 연산을 줄이는 방법을 제안한다. CAM shift는 대표적인 색상을 이용한 추적 방법이나 빠른 속도로 이동하는 물체를 추적하기 어려운 단점이 있다. 이를 해결하기 위해 추적 대상을 놓쳐버린 시점에서 마지막으로 추적에 성공한 시점의 정보를 이용하여 8방향 탐색을 실시하여 객체를 찾아 낸 후 CAM shift의 탐색 윈도우를 이동시켜 기존의 CAM shift로는 추적이 불가능한 고속 이동 물체에 대해서도 보다 정확한 추적이 가능하게 되었다. 또한 하드웨어의 발달로 초당 생산되어지는 프레임의 수가 증가하여 불필요한 연산이 증가하게 되었고, 이를 줄이기 위해 추적에 이용되는 프레임의 수를 줄여 연산을 줄여 이 전보다 효율을 높일 수 있었다.

This research aims to suggest methods to improve object tracking performance by combining CAM shift algorithm with 8-way search window, and reduce arithmetic operation by reducing the number of frame used for tracking. CAM shift has its adverse effect in tracking methods using signature color or having difficulty in tracking rapidly moving object. To resolve this, moving search window of CAM shift makes it possible to more accurately track high-speed moving object after finding object by conducting 8-way search by using information at a final successful timing point from a timing point missing tracking object. Moreover, hardware development led to increased unnecessary arithmetic operation by increasing the number of frame produced per second, which indicates efficiency can be enhanced by reducing the number of frame used in tracking to reduce unnecessary arithmetic operation.