Video captioning is a challenging task as it needs to accurately transform visual understanding into natural language description. To date, state-of-the-art methods inadequately model global-local vision representation for sentence generation, leaving plenty of room improvement. In this work, we approach the video from new perspective and propose GLR framework, namely granularity. Our demonstrates three advantages over prior efforts. First, simple solution, which exploits extensive...

Optical flow is an indispensable building block for various important computer vision tasks, including motion estimation, object tracking, and disparity measurement. In this work, we propose TransFlow, a pure transformer architecture optical estimation. Compared to dominant CNN-based methods, TransFlow demonstrates three advantages. First, it provides more accurate correlation trustworthy matching in estimation by utilizing spatial self-attention crossattention mechanisms between adjacent...

Attribute-based encryption (ABE) with outsourced decryption not only enables fine-grained sharing of encrypted data, but also overcomes the efficiency drawback (in terms ciphertext size and cost) standard ABE schemes. In particular, an scheme allows a third party (e.g., cloud server) to transform into (short) El Gamal-type using public transformation key provided by user so that latter can be decrypted much more efficiently than former user. However, shortcoming original is correctness...

As cloud computing becomes prevalent, more and sensitive data is being centralized into the by users. To maintain confidentiality of user against untrusted servers, should be encrypted before they are uploaded. However, this raises a new challenge for performing search over efficiently. Although existing searchable encryption schemes allow to with confidentiality, these solutions cannot support verifiability searching result. We argue that server may selfish in order save its computation...

In this paper, we propose a lightweight privacy-preserving Faster R-CNN framework (SecRCNN) for object detection in medical images. is one of the most outstanding deep learning models detection. Using SecRCNN, healthcare centers can efficiently complete computations via additive secret sharing technique and edge computing. To implement design series interactive protocols to perform three stages R-CNN, namely feature map extraction, region proposal regression classification. improve...

Abstract Applying intelligence algorithms to conceive nanoscale meta-devices becomes a flourishing and extremely active scientific topic over the past few years. Inverse design of functional nanostructures is at heart this topic, in which artificial (AI) furnishes various optimization toolboxes speed up prototyping photonic layouts with enhanced performance. In review, we offer systemic view on recent advancements nanophotonic components designed by algorithms, manifesting development trend...

Instance segmentation in videos, which aims to segment and track multiple objects video frames, has garnered a flurry of research attention recent years. In this paper, we present novel weakly supervised framework with \textbf{S}patio-\textbf{T}emporal \textbf{C}ollaboration for instance \textbf{Seg}mentation namely \textbf{STC-Seg}. Concretely, STC-Seg demonstrates four contributions. First, leverage the complementary representations from unsupervised depth estimation optical flow produce...

Over time, natural materials have evolved to be lightweight, high-strength, tough, and damage-tolerant due their unique biological structures. Therefore, combining inspiration structural design would provide traditional with a broader range of performance applications. Here, the application an ink-based three-dimensional (3D) printing strategy Lunar regolith simulant-based geopolymer (HIT-LRS-1 GP) was first reported, high-precision carbon fiber/quartz sand-reinforced biomimetic patterns...

The immutable feature of blockchain determines that traditional security response mechanisms (e.g., code patching) must change to remedy insecure smart contracts. only proper way protect a contract is fix potential risks in its before it deployed the blockchain. However, existing tools for analysis focus on detection bugs but seldom consider issues. Meanwhile, often time-consuming and error-prone developer understand flawed manually. In this paper we propose SMARTSHIELD, bytecode...

Recently, Google and other 24 institutions proposed a series of open challenges towards federated learning (FL), which include application expansion homomorphic encryption (HE). The former aims to expand the applicable machine models FL. latter focuses on who holds secret key when applying HE For naive scheme, server is set master key. Such setting causes serious problem that if does not conduct aggregation before decryption, chance left for access user's update. Inspired by two challenges,...

Cryptography is increasingly being used in mobile applications to provide various security services; from user authentication, data privacy, secure communications. However, there are plenty of mistakes that developers could accidentally make when using cryptography their apps and such can lead a false sense security. Recent research efforts indeed show significant portion both Android iOS platforms misused cryptographic APIs. In this paper, we present CDRep, tool for automatically repairing...

Smart contracts, programs running on blockchain systems, leverage diverse decentralized applications (DApps). Unfortunately, well-known smart contract platforms, Ethereum for example, face serious security problems. Exploits to contracts may cause enormous financial losses, which emphasize the importance of testing. However, current exploit generation tools have difficulty solve hard constraints in execution paths and cannot simulate behaviors very well. These problems a loss coverage...

Recently, Google and other 24 institutions proposed a series of open challenges towards federated learning (FL), which include application expansion homomorphic encryption (HE). The former aims to expand the applicable machine models FL. latter focuses on who holds secret key when applying HE For naive scheme, server is set master key. Such setting causes serious problem that if does not conduct aggregation before decryption, chance left for access user's update. Inspired by two challenges,...

Nowadays, the pseudonym swap has become mainstream technology for protecting vehicles' trajectory privacy in vehicle ad hoc networks. However, existing methods cannot strictly provide unlinkability between new and old of due to lack theoretical guarantee, resulting severe leakages privacy. Our experiment also proves this point we find that works may cause vehicle's pseudonyms be linked with a probability higher than 60% because they always choose two vehicles very different driving states...

Searchable encryption(SE) allows users to efficiently retrieve data over encrypted cloud data, but most existing SE schemes only support exact keyword search, resulting in false results due minor typos or format inconsistencies of queried keywords. The fuzzy search can avoid this limitation, still incurs low accuracy and efficiency. Besides, do not consider malicious servers which may execute a fraction operations forge some various interest incentives such as saving computation storage...

Abstract In this study, we report that geopolymer materials containing different amounts of graphitic carbon nitride (g‐C 3 N 4 ) (C /GP) can remove methylene blue (MB) using the synergistic effects adsorption and photodegradation. A wt% g‐C (3 wt%C had best MB removal efficiency 94.62% within 60 min, under visible light. The results showed geopolymers act as carriers for , relative amount adsorbent photocatalyst influences MB. /GP was also used additive manufacturing by direct ink writing...

Developers and manufacturers provide configurable control parameters for flight programs to support various environments missions, along with suggested ranges these ensure safety. However, this flexible mechanism can also introduce a vulnerability known as range specification bugs. The originates from the evidence that certain combinations of parameter values may affect drone's physical stability even though its are within range. paper introduces novel system called <small>icsearcher</small>...

In this study, new metal-organic frameworks (MOFs) nanocrystals modified SiO2 core-shell microspheres were designed with cationic ionic liquids (ILs) 1,3-bis(4-carboxybutyl)imidazolium bromide (ILI) as organic ligands. By further adjustment the growth cycles, ILI-01@SiO2 stationary phase was facilely fabricated. The developed respectively characterized via element analysis, thermogravimetric scanning electron microscopy, X-ray diffraction, and Fourier transform infrared spectrometry. Because...