Homomorphic Encryption (HE) based secure Neural Networks(NNs) inference is one of the most promising security solutions to emerging Machine Learning as a Service (MLaaS). In the HE-based MLaaS setting, a client encrypts the sensitive data, and uploads the encrypted data to the server that directly processes the encrypted data without decryption, and returns the encrypted result to the client. The clients' data privacy is preserved since only the client has the private key. Existing HE-enabled Neural Networks (HENNs), however, suffer from heavy computational overheads. The state-of-the-art HENNs adopt ciphertext packing techniques to reduce homomorphic multiplications by packing multiple messages into one single ciphertext.

The client'S data privacy is preserved since only the client has the private key. Existing HE-enabled Neural Networks (HENNs), however, suffer from heavy computational overheads.

Upon arrival at Japan's Henn na Hotel, you are greeted by a pair of receptionists nodding from behind the front desk as you check in at a tablet. A quiet grace emanates from their serene smiles, confident gaze, and perfect porcelain skin. Say "good evening," and they may blink. Ask for the weather report, and they may reply, "Tomorrow's weather is fine and 25C." They wear pristine white uniforms, blue silk scarves, and white caps that sit perfectly atop their glossy black bobs.

Homomorphic Encryption (HE) based secure Neural Networks(NNs) inference is one of the most promising security solutions to emerging Machine Learning as a Service (MLaaS). In the HE-based MLaaS setting, a client encrypts the sensitive data, and uploads the encrypted data to the server that directly processes the encrypted data without decryption, and returns the encrypted result to the client. The clients' data privacy is preserved since only the client has the private key. Existing HE-enabled Neural Networks (HENNs), however, suffer from heavy computational overheads. The state-of-the-art HENNs adopt ciphertext packing techniques to reduce homomorphic multiplications by packing multiple messages into one single ciphertext.

Deep neural networks (DNNs) have been shown to perform well on exclusive, multi-class classification tasks. However, when different classes have similar visual features, it becomes challenging for human annotators to differentiate them. This scenario necessitates the use of composite class labels. In this paper, we propose a novel framework called Hyper-Evidential Neural Network (HENN) that explicitly models predictive uncertainty due to composite class labels in training data in the context of the belief theory called Subjective Logic (SL). By placing a grouped Dirichlet distribution on the class probabilities, we treat predictions of a neural network as parameters of hyper-subjective opinions and learn the network that collects both single and composite evidence leading to these hyper-opinions by a deterministic DNN from data. We introduce a new uncertainty type called vagueness originally designed for hyper-opinions in SL to quantify composite classification uncertainty for DNNs. Our results demonstrate that HENN outperforms its state-of-the-art counterparts based on four image datasets. The code and datasets are available at: https://github.com/Hugo101/HyperEvidentialNN.

I met my first South Korean robots as I checked into the Henn na Hotel in Seoul at the end of a 21-hour journey from the U.S.: two plane flights and a bleary-eyed ride on the transit rail. Behind the front desk stood two gleaming white androids, with big round heads framing green digital eyes and thin green smiles. I headed for the androids. The robot clerk on the right came alive to greet me -- first in English, then in Korean, Japanese, and Chinese, in quick succession. "Welcome to the Henn na Hotel!" it said in a chirpy female voice. It was eerily humanoid yet inhuman, with hands that looked like white-fingered gloves and thin black mechanical joints for elbows. Its cartoony face was drawn for friendliness. Its slender arms occasionally swept outward in a welcoming gesture.

Internet-of things is seemly always vulnerable to security flaws. From individual users to the corporate sector, these IoT flaws have always impacted users. Once again, a Japanese hotel fell victim to such a vulnerability in its in-room robots. Exploiting the flaw could allow spying on the customers. Security researcher Lance R. Vick spotted a vulnerability in the Tapia robots installed in a Japanese hotel.

Many of the unusual robots working at Japan's Henn na Hotel are now out of work. Humans worried about the robot revolution threatening their jobs can relax. Japan's Henn na Hotel prided itself on its all-robot staff, but it turns out they weren't up to the job. Henn na's robot staff was first employed in 2015 with the aim of becoming "the most efficient hotel in the world." But four years later, it seems that the 243 robots are less of a novelty and more of a nuisance.

A hotel chain in Tokyo is trying to be the first run by robots, albeit with a few malfunctions along the way, according to a new report. Japanese travel company H.I.S. has been opening hotels where robots man the front desk, check-in is handled by a kiosk, and face recognition opens the door to your room. The hotel name, Henn na Hotel ("henn na" means strange in Japanese) underscores the eerie presence of robots and the relative dearth of humans. The first hotel opened in 2015 and was recognized by Guinness World Records as "the first robot-staffed hotel" in the world, according to the Japan Times. H.I.S. is slowly expanding the number of robot-centric locations.