科研成果 by Type: Conference Paper

This paper provides the first piece of empirical evidence regarding the impact of health cost risk on individuals' annuitization decisions. We find that health cost risk increases the probability of individuals' pension participation but decreases the amount of pension contributions. We show that the substitution effect of informal insurance on pensions leads to these seemingly contradictory results. The impact of health cost risk on pension participation and contributions is negative and consistent with the mainstream theory after accounting for the effect of informal insurance. The substitution effect of informal insurance on pensions is stronger, and thus mitigates the impact of health cost risk more pronounced for households that have better-educated children, lower incomes, and more informal social networks and in regions that have a higher male–female ratio, that have higher mobility, or are less developed; but this substitution effect does not differ depending on their children's gender. This study improves our understanding of the relationship between health cost risk and individuals' annuitization decisions as well as the role of informal insurance in this relationship.

Multi-focus image fusion (MFIF) is a critical technique for enhancing depth of field in photography, producing an all-in-focus image from multiple images captured at different focal lengths. While deep learning has shown promise in MFIF, most existing methods ignore the physical model of defocus blurring in their neural architecture design, limiting their interoperability and generalization. This paper presents a novel framework that integrates explicit defocus blur modeling into the MFIF process, leading to enhanced interpretability and performance. Leveraging an atom-based spatially-varying parameterized defocus blurring model, our approach first calculates pixel-wise defocus descriptors and initial focused images from multi-focus source images through a scale-recurrent fashion, based on which soft decision maps are estimated. Afterward, image fusion is performed using masks constructed from the decision maps, with a separate treatment on pixels that are probably defocused in all source images or near boundaries of defocused/focused regions. Model training is done with a fusion loss and a cross-scale defocus estimation loss. Extensive experiments on benchmark datasets have demonstrated the effectiveness of our approach.

A few colonies in the Latin West (Augusta Emerita, Johnston 2017: 56-57 fn. 202-203; Abascal 2002: 284) and the Greek East (e.g. Philippi AE 1932, 21; briefly Samos IG XII,6 1:187 l. 8-11) used the annus coloniae, the year when the colony was founded, as their standard time-reckoning formula on inscriptions and perhaps other media as well. The list may include the Roman colony of Sinope, which coin issues carried legends such as C(olonia) F(elix) I(ulia) an(no) followed by (and later on simply with) Latin numerals (e.g. RPC I 2110, II 715). Curiously, since the 17th century, numismatists have observed two epochs on Sinope’s Roman period issues (e.g. Hardouin 1689: 140). There was the epoch of 45 BCE, likely the annus coloniae (e.g. RPC III 1229), as it coincided with reports of a Caesarian colony (Strab. 12.3.11). Then there was the epoch of 70 BCE (e.g. RPC VI 6501), seemingly replacing the earlier epoch, but only from the early third century onwards (Kubitschek 1908: 68-71; Magie 1950: 342 no. 42).  Changing epochs was not a rare phenomenon in Anatolia, but known examples changed an earlier epoch for a later one, apparently to adapt to new circumstances, such as an imperial victory or visit (Leschhorn 1993: 439-541), and count years from a new beginning. Sinope, however, changed the epoch that was supposedly linked to the year of its colonial foundation to an earlier epoch that coincided with Lucullus’ “liberation” of Sinope from Mithridatid control. This suggests that Sinope was adapting to new circumstances that required the adaptation of an historical epoch. It is not the beginning per se, but a recaliberation of municipal history. While historical developments of Roman Sinope is not well documented in extant literary sources, and epigraphic evidence is scarce, this paper wishes to take two approaches to assess the problem of Sinope’s changing epochs. The first approach is to consider Sinope’s time-reckoning tradition. The era system that was prevalent in northern Anatolia and elsewhere during the first three centuries CE had a strong Hellenistic tradition. In this view, there may be more nuance – 45 BCE can be the annus coloniae, but it was reflecting a recent change in the city’s history, not an overhaul of the city’s established time-reckoning system, perhaps based on the Seleucid era (found on ceramic stamps, Saprykin & Fedoseev 1999: 135-143). This approach removes the Roman symbolism attached to the 45 BCE, and opens up additional possibilities for interpretation. The second approach considers the epoch of 70 BCE regionally. This epoch was primarily used at Amastris, but also attested at Abonuteichos. Both cities were described by Roman authors as prosperous in their own ways (e.g. Luc. Alex. 9, 25; Plin. Ep. 10.98). Both cities also began to issue 3 to 4 units of local bronze coinage from the reign of Trajan onwards, including the 6-assaria, associated with economic prosperity (Zajac 2023: 30-32, Tab. 1a). Epigraphic and numismatic evidence further suggest that Amastrians were mobile across the Black Sea and the Aegean, some even serving as koinon officials of Bithynian and other regional koina (Marek 2003: 63-67). There is the possibility that Sinope changed its epoch to adhere to regional time-reckoning norms set by more dominant peers. Since the 17th century, numismatists have observed two epochs on Sinope’s Roman period coin issues (e.g. Hardouin 1689: 140). There was the epoch of 45 BCE, perhaps the annus coloniae (e.g. RPC III 1229; cf. Augusta Emerita, Johnston 2017: 56-57 fn. 202-203; Abascal 2002: 284; Philippi AE 1932, 21; briefly Samos IG XII,6 1:187 l. 8-11), as it coincided with reports of a Caesarian colony (Strab. 12.3.11). Then there was the epoch of 70 BCE (e.g. RPC VI 6501), seemingly replacing the earlier epoch, but only from the early third century onwards (Kubitschek 1908: 68-71; Magie 1950: 342 no. 42). Neither changing epochs or the use of many epochs were rare phenomena. That said, known examples show cities replacing old epochs with those that commemorate new events and circumstances (Leschhorn 1993: 439-541). It is then odd for Sinope to abandon the epoch commemorating its colonial foundation in favor of an epoch 25 years earlier. Why? Leschhorn gave two potential scenarios: perhaps 70 BCE was the outcome a pro-Severan party defeating the old guard, or there was a “Gräzisierung" movement under which Sinope opted for a symbolically more acceptable epoch (Leschhorn 1993: 162), but these suggestions are limited by lack of evidence. This paper reassesses Sinope’s changing epochs with two approaches. First is a reframing of the Leschhorn’s assumed constitutional symbolism attached to the epoch of 45 BCE that leads to factionalism, and the “Roman” nature of Sinope’s so-called Kolonieära in Leschhorn’s alternative “Gräzisierung” thesis. The question asked here is whether Sinope had an established era system in place as part of its Hellenistic past. Sinope likely used the Pontic era while under Mithridatid control (Leschhorn 1993: 150). Recent discussions of stamped amphorae further indicates that the Seleucid era system was likely present if not also used at Sinope (Saprykin & Fedoseev 1999: 135-143). Further, we now have ceramic tiles time-stamped with a formula such as "s(aeculo) S(inopensis) (colonia) i(nducta) an(no) IV" (Fedoseev 2019: 16-17). If Fedoseev’s reading of the time-stamp is correct, we may consider the possibility that the saeculum Sinopensis was a system that could have accommodated multiple epochs, and the phenomenon in question was a symptom of an entrenched but adaptative time-reckoning mechanism at work. The second approach considers the epoch of 70 BCE regionally. This epoch was primarily used at Amastris, but also attested at Abonuteichos. Both cities were described by Roman authors as prosperous in their own ways (e.g. Luc. Alex. 9, 25; Plin. Ep. 10.98). Both cities also began to issue 3 to 4 units of local bronze coinage from the reign of Trajan onwards, including the 6-assaria, associated with economic prosperity (Zajac 2023: 30-32, Tab. 1a). Epigraphic and numismatic evidence further suggest that Amastrians were mobile across the Black Sea and the Aegean, some even serving as koinon officials of Bithynian and other regional koina (Marek 2003: 63-67). There is the possibility that Sinope changed its epoch to adhere to regional time-reckoning norms set by more dominant peers.

Low-light image enhancement (LLIE) aims to improve visibility and signal-to-noise ratio in images captured under poor lighting conditions. Despite impressive improvement, deep learning-based LLIE approaches require extensive training data, which is often difficult and costly to obtain. In this paper, we propose a zero-shot LLIE framework leveraging pre-trained latent diffusion models for the first time, which act as powerful priors to recover latent images from low-light inputs. Our approach introduces several components to alleviate the inherent challenges in utilizing pre-trained latent diffusion models, modeling the degradation process in an image-adaptive manner, penalizing the latent outside the manifold of natural images, and balancing the strengths of the guidance from the given low-light image during the denoising process. Experimental results demonstrate that our framework outperforms existing methods, achieving superior performance across various datasets.

Hyperspectral imaging plays a critical role in numerous scientific and industrial fields. Conventional hyperspectral imaging systems often struggle with the trade-off between spectral and temporal resolution, particularly in dynamic environments. In ours work, we present an innovative event-based active hyperspectral imaging system designed for real-time performance in dynamic scenes. By integrating a diffraction grating and rotating mirror with an event-based camera, the proposed system captures high-fidelity spectral information at a microsecond temporal resolution, leveraging the event camera's unique capability to detect instantaneous changes in brightness rather than absolute intensity. The proposed system trade-off between conventional frame-based systems by reducing the bandwidth and computational load and mosaic-based system by remaining the original sensor spatial resolution. It records only meaningful changes in brightness, achieving high temporal and spectral resolution with minimal latency and is practical for real-time applications in complex dynamic conditions.

Automotive audio systems often face sub-optimal sound quality due to the intricate acoustic properties of car cabins. Acoustic channel equalization methods are generally employed to improve sound reproduction quality in such environments. In this paper, we propose an acoustic channel equalization method using convex optimization in the modal domain. The modal domain representation is used to model the whole sound field to be equalized. Besides integrating it into the convex formulation of the acoustic channel reshaping problem, to further control the prering artifacts, the temporal window function modified according to the backward masking effect of the human auditory system is used during equalizer design. Objective and subjective experiments in a real automotive cabin proved that the proposed method enhances spatial robustness and avoids the audible prering artifacts.