Search for: All records

Navigating safely and independently presents considerable challenges for people who are blind or have low vision (BLV), as it re- quires a comprehensive understanding of their neighborhood environments. Our user study reveals that understanding sidewalk materials and objects on the sidewalks plays a crucial role in navigation tasks. This paper presents a pioneering study in the field of navigational aids for BLV individuals. We investigate the feasibility of using auditory data, specifically the sounds produced by cane tips against various sidewalk materials, to achieve material identification. Our approach utilizes ma- chine learning and deep learning techniques to classify sidewalk materials solely based on audio cues, marking a significant step towards empowering BLV individuals with greater autonomy in their navigation. This study contributes in two major ways: Firstly, a lightweight and practical method is developed for volunteers or BLV individuals to autonomously collect auditory data of sidewalk materials using a microphone-equipped white cane. This innovative approach transforms routine cane usage into an effective data-collection tool. Secondly, a deep learning-based classifier algorithm is designed that leverages a dual architecture to enhance audio feature extraction. This includes a pre-trained Convolutional Neural Network (CNN) for regional feature extraction from two-dimensional Mel-spectrograms and a booster module for global feature enrichment. Experimental results indicate that the optimal model achieves an accuracy of 80.96% using audio data only, which can effectively recognize sidewalk materials. 

Navigating safely and independently presents considerable challenges for people who are blind or have low vision (BLV), as it re- quires a comprehensive understanding of their neighborhood environments. Our user study reveals that understanding sidewalk materials and objects on the sidewalks plays a crucial role in navigation tasks. This paper presents a pioneering study in the field of navigational aids for BLV individuals. We investigate the feasibility of using auditory data, specifically the sounds produced by cane tips against various sidewalk materials, to achieve material identification. Our approach utilizes ma- chine learning and deep learning techniques to classify sidewalk materials solely based on audio cues, marking a significant step towards empowering BLV individuals with greater autonomy in their navigation. This study contributes in two major ways: Firstly, a lightweight and practical method is developed for volunteers or BLV individuals to autonomously collect auditory data of sidewalk materials using a microphone-equipped white cane. This innovative approach transforms routine cane usage into an effective data-collection tool. Secondly, a deep learning-based classifier algorithm is designed that leverages a dual architecture to enhance audio feature extraction. This includes a pre-trained Convolutional Neural Network (CNN) for regional feature extraction from two-dimensional Mel-spectrograms and a booster module for global feature enrichment. 

Context.An accurate²⁸P(p,γ)²⁹S reaction rate is crucial to defining the nucleosynthesis products of explosive hydrogen burning in ONe novae. Using the recently released nuclear mass of²⁹S, together with a shell model and a direct capture calculation, we reanalyzed the²⁸P(p,γ)²⁹S thermonuclear reaction rate and its astrophysical implication. Aims.We focus on improving the astrophysical rate for²⁸P(p,γ)²⁹S based on the newest nuclear mass data. Our goal is to explore the impact of the new rate and associated uncertainties on the nova nucleosynthesis. Methods.We evaluated this reaction rate via the sum of the isolated resonance contribution instead of the previously used Hauser-Feshbach statistical model. The corresponding rate uncertainty at different energies was derived using a Monte Carlo method. Nova nucleosynthesis is computed with the 1D hydrodynamic code SHIVA. Results.The contribution from the capture on the first excited state at 105.64 keV in²⁸P is taken into account for the first time. We find that the capture rate on the first excited state in²⁸P is up to more than 12 times larger than the ground-state capture rate in the temperature region of 2.5 × 10⁷K to 4 × 10⁸K, resulting in the total²⁸P(p,γ)²⁹S reaction rate being enhanced by a factor of up to 1.4 at ~1 × 10⁹K. In addition, the rate uncertainty has been quantified for the first time. It is found that the new rate is smaller than the previous statistical model rates, but it still agrees with them within uncertainties for nova temperatures. The statistical model appears to be roughly valid for the rate estimation of this reaction in the nova nucleosynthesis scenario. Using the 1D hydrodynamic code SHIVA, we performed the nucleosynthesis calculations in a nova explosion to investigate the impact of the new rates of²⁸P(p,γ)²⁹S. Our calculations show that the nova abundance pattern is only marginally affected if we use our new rates with respect to the same simulations but statistical model rates. Finally, the isotopes whose abundance is most influenced by the present²⁸P(p,γ)²⁹S uncertainty are²⁸Si,^33,34S,^35,37Cl, and³⁶Ar, with relative abundance changes at the level of only 3% to 4%.