This paper develops a bilateral symmetric trade model with firms engaging in strategic exporting behaviour. It provides a unified analysis of trade liberalisation allowing for a rich set of cost structures both on the production side and the transportation side that go beyond the standard assumption of constant marginal cost. The theoretical analysis reveals among others that trade liberalisation does not always result in welfare gains. In fact, trade liberalisation coinciding with a reduction in the transport cost can significantly reduce consumer welfare. Depending on the transport cost and the fixed cost of exporting, almost all possible trade patterns can appear. Trade liberalisation may even prevent a country from participating in exports when fixed costs are relatively high. We thus provide a potential theoretical rationale for the many existing empirical studies that report results that diverge from standard theories, by bringing it back to the diverse cost structures that may occur in the real world beyond the ones assumed in the existing theories.
In recent years, detecting logical anomalies that occur when objects deviate from rules regarding their quantities or spatial arrangements has been the focus of research. This study proposes a method for detecting anomalies in the types and numbers of parts arranged across multiple partitions without explicitly pre-setting rules for normal products. The proposed method automatically extracts a normal distribution representing the common rules of normal products across partitions by using features extracted from each region segment, and it judges deviations from this normal distribution as anomalies. Experiments conducted using MVTec LOCO AD demonstrated that the proposed method achieved an AUROC of 93.6%, surpassing ComAD (89.6%) and EfficientAD (71.6%), both of which do not require prior information about the rules of normal products, and showing performance equal to or higher than state-of-the-art methods that require explicit prior information about such rules, including CSAD (92.6%) and PSAD (93.6%).
In this study, we propose a system that enables a robot to generate a sequence of actions for real-world cooking by leveraging a large language model (LLM) together with simple recipe texts-such as "Boil the egg for 5 minutes"-and visual information captured by the robot's camera. The generated action sequence is evaluated based on a set of custom-designed criteria, and it is corrected as necessary. In our experiments, the proposed system achieved a sequence generation success rate of 88% across three types of cooking actions. Furthermore, a robot execution success rate of 73% was confirmed for two of the cooking actions, demonstrating the effectiveness of the proposed system.
This paper proposes a method for predicting console time of laparoscopic partial nephrectomy based on radiomics features extracted from the perinephric fat region on CT images. We constructed a LightGBM-based regressor using 109 radiomics features and 30 clinical variables, and demonstrated that the model incorporating image features reduced prediction error by up to 22 minutes compared with a model using only clinical information.
The association between brand name sounds and taste perception is an emerging area of interest in marketing research. This study aims to demonstrate the role of sound-evoked cuteness in the expectation of sweet taste. Across seven studies (including two supplementary studies), our findings revealed that sound reduplication in brand names is consistently associated with sweet taste expectations. This effect was mediated by perceptions of baby schema cuteness. Furthermore, our results revealed that sweet-liker status and brand status moderated the effects of sound reduplication. Brands with sound reduplication (vs. non-reduplication) were more preferred by sweet-likers than by sweet-dislikers. Sound reduplication (vs. non-reduplication) also increased preferences for mass-market brands (vs. luxury brands). Our findings identify the reduplication of brand name sounds as a novel antecedent in the context of food branding.
