Improved Spoken Language Representation for Intent Understanding in a Task-Oriented Dialogue System.

Successful applications of deep learning technologies in the natural language processing domain have improved text-based intent classifications. However, in practical spoken dialogue applications, the users' articulation styles and background noises cause automatic speech recognition (ASR) errors, and these may lead language models to misclassify users' intents. To overcome the limited performance of the intent classification task in the spoken dialogue system, we propose a novel approach that jointly uses both recognized text obtained by the ASR model and a given labeled text. In the evaluation phase, only the fine-tuned recognized language model (RLM) is used. The experimental results show that the proposed scheme is effective at classifying intents in the spoken dialogue system containing ASR errors.

A Military Audio Dataset for Situational Awareness and Surveillance.

Audio classification related to military activities is a challenging task due to the high levels of background noise and the lack of suitable and publicly available datasets. To bridge this gap, this paper constructs and introduces a new military audio dataset, named MAD, which is suitable for training and evaluating audio classification systems. The proposed MAD dataset is extracted from various military videos and contains 8,075 sound samples from 7 classes corresponding to approximately 12 hours, exhibiting distinctive characteristics not presented in academic datasets typically used for machine learning research. We present a comprehensive description of the dataset, including its acoustic statistics and examples. We further conduct a comprehensive sound classification study of various deep learning algorithms on the MAD dataset. We are also releasing the source code to make it easy to build these systems. The presented dataset will be a valuable resource for evaluating the performance of existing algorithms and for advancing research in the field of acoustic-based hazardous situation surveillance systems.

Bowen's disease concealed by purpura.

Bowen's disease (BD) is a squamous cell carcinoma in situ characterized by a well-demarcated scaly erythematous thin plaque with an irregular outline. Clinically, BD is frequently misdiagnosed as superficial basal cell carcinoma, patches of dermatitis, psoriasis, lichen planus, actinic keratosis, benign lichenoid keratosis, irritated seborrheic keratosis, viral warts, amelanotic melanoma or melanoma. However, angiosarcoma has not usually been mentioned in the differential diagnosis of BD before. Herein, we describe two cases of BD presenting as purpura on the scalp of the elderly with an initial clinical suspicion of angiosarcoma.

Fibrous hamartoma of infancy manifesting as multiple nodules with hypertrichosis.

Fibrous hamartoma of infancy is a rare benign tumor, which usually occurs within the first year of life. Most cases present as a solitary, painless, skin-colored subcutaneous nodule; multiple lesions are rarely reported. There have also been a few cases which have showed overlying skin changes, including hypertrichosis and alterations in pigmentation. We report a rare case of fibrous hamartoma of infancy in a 6-month-old female who presented as multiple nodules with overlying hypertrichosis on the left buttock. To the best of our knowledge, there have been only two reported cases in the dermatological published work that showed these unusual findings simultaneously.

Two cases of perforating lichen nitidus.

Lichen nitidus is an unusual condition that usually affects children and young adults. Of its many variants, a perforating type has been rarely reported. We herein report two cases of lichen nitidus with perforation. Two young women in good health presented with multiple shiny papules that seems to be typical lichen nitidus clinically, but some of the papules were umbilicated. Histopathological examination revealed the typical findings of lichen nitidus with perforation in both cases.