Deep Neural Networks (DNN) are gaining higher commercial values in computer vision applications, e.g., image classification, video analytics, etc. This calls for urgent demands of the intellectual ...property (IP) protection of DNN models. In this paper, we present a novel watermarking scheme to achieve the ownership verification of DNN architectures. Existing works all embedded watermarks into the model parameters while treating the architecture as public property. These solutions were proven to be vulnerable by an adversary to detect or remove the watermarks. In contrast, we claim the model architectures as an important IP for model owners, and propose to implant watermarks into the architectures. We design new algorithms based on Neural Architecture Search (NAS) to generate watermarked architectures, which are unique enough to represent the ownership, while maintaining high model usability. Such watermarks can be extracted via side-channel-based model extraction techniques with high fidelity. We conduct comprehensive experiments on watermarked CNN models for image classification tasks and the experimental results show our scheme has negligible impact on the model performance, and exhibits strong robustness against various model transformations and adaptive attacks.
Purpose
Proper taste-masking formulation design is a critical issue for instant-dissolving tablets (IDTs). The purpose of this study is to use the electronic tongue to design the additives of the 3D ...printed IDTs to improve palatability.
Methods
A binder jet 3D printer was used to prepare IDTs of levetiracetam. A texture analyzer and dissolution apparatus were used to predict the oral dispersion time and
in vitro
drug release of IDTs, respectively. The palatability of different formulations was investigated using the ASTREE electronic tongue in combination with the design of experiment and a model for masking bitter taste. Human gustatory sensation tests were conducted to further evaluate the credibility of the results.
Results
The 3D printed tablets exhibited rapid dispersion (<30 s) and drug release (2.5 min > 90%). The electronic tongue had an excellent ability of taste discrimination, and levetiracetam had a good linear sensing performance based on a partial least square regression analysis. The principal component analysis was used to analyze the signal intensities of different formulations and showed that 2% sucralose and 0.5% spearmint flavoring masked the bitterness well and resembled the taste of corresponding placebo. The results of human gustatory sensation test were consistent with the trend of the electronic tongue evaluation.
Conclusions
Owing to its objectivity and reproducibility, this technique is suitable for the design and evaluation of palatability in 3D printed IDT development.
The main problem for the application of hydrogen generated via hydrolysis of metal alloys is the low hydrogen generation rate (HGR). In this paper, active Al alloys were prepared using a new coupled ...method-melting-mechanical crushing-mechanical ball milling method to enhance the HGR at room temperature. This method contains three steps, including the melting of Al, Ga, In, and Sn ingots with low melting alloy blocks and casting into plates, then crushing alloy plate into powders and ball milling with chloride salts such as NiCl
and CoCl
were added during the ball milling process. The microstructure and phase compositions of Al alloys and reaction products were investigated via X-ray diffraction and scanning electron microscopy with energy dispersed X-ray spectroscopy. The low-melting-point Ga-In -Sn (GIS) phases contain a large amount of Al can act as a transmission medium for Al, which improves the diffusion of Al to Al/H
O reaction sites. Finer GIS phases after ball milling can further enhance the diffusion of Al and thus enhance the activity of Al alloy. The hydrogen generation performance through hydrolysis of water with Al at room temperature was investigated. The results show that the H
generation performance of the Al-low-melting point alloy composite powder is significantly higher than the results reported to date. The highest H
generation rate and H
conversion efficiency can reach 5337 mL·min
·g
for the hydrolysis of water with 1 g active alloy.
For exploring an effective heat treatment schedule to enhance the strength-plasticity balance of the ferrite-austenite 12CrNi2 alloy steel additively manufactured by directed energy deposition (DED), ...12CrNi2 was heat-treated with deliberately designed direct quenching (DQ) and cyclic quenching (CQ), respectively, and the differently quenched steels were then tempered at a temperature from 200 °C to 500 °C. It was found that the CQ, in contrast to the DQ, led the 12CrNi2 to have significantly increased tensile strength without losing its plasticity, based on the introduction of fine-grained lath martensite and the {112} -type nanotwins. The nanotwins were completely degenerated after the 200 °C tempering. This led the CQ-treated steel to decrease in not only its tensile strength, but also its plasticity. In addition, an interesting phenomenon observed was that the DQ-induced laths and rod-like precipitates, and the tempering-induced laths and rod-like precipitates were all prone to be generated along the {112} planes of the martensitic crystal (α-Fe), which were exactly fitted with the {112}-type crystalline orientation of the long or short nanotwins in the CQ-induced martensite. The quenching-tempering-induced generation of the {112}-orientated laths and rod-like precipitates was explicated in connection with the {112} -type long or short nanotwins in the CQ-induced lath martensite.
The limited lymphocytes infiltration and immunosuppression in tumor are the major challenges of cancer immunotherapy. The use of immunogenic cell death (ICD)‐inducing agents has potential to ...potentiate antitumor immune responses, but is tremendously hampered by the poor delivery efficiency. Herein, a tumor‐activated size‐enlargeable bioinspired lipoprotein of oxaliplatin (TA‐OBL) is designed to access cancer cells and boost the ICD‐induced antitumor immunity for synergizing immune‐checkpoint blockades (ICBs)‐mediated immunotherapy. TA‐OBL is constructed by integrating a legumain‐sensitive melittin conjugate for improving intratumoral permeation and cancer cell accessibility, a pH‐sensitive phospholipid for triggering size‐enlargement and drug release in intracellular acidic environments, a nitroreductase‐sensitive hydrophobic oxaliplatin prodrug (N‐OXP) for eliciting antitumor immunity into the bioinspired nano‐sized lipoprotein system. TA‐OBL treatment produced robust antitumor immune responses and its combination with ICBs demonstrates strong therapeutic benefits with delayed tumor growth and extended survival rate, making it a promising delivery nanoplatform to elicit antitumor immunity for cancer immunotherapy.
The use of immunogenic cell death (ICD)‐inducing agents has potential to potentiate antitumor immune responses and remodel the immunosuppression in tumor, but is tremendously hampered by the poor delivery efficiency. A tumor‐activated size‐enlargeable bioinspired lipoprotein of oxaliplatin prodrug (TA‐OBL) with efficient cancer cell accessibility is designed to boost the ICD‐induced antitumor immunity for synergizing immune‐checkpoint‐blockades‐mediated cancer immunotherapy.
To investigate the direct influence and mechanism of China's financial and manufacturing co-agglomeration on environmental pollution, we constructed a panel data regression model incorporating ...mediating and threshold effects with the panel data of 285 prefecture-level cities from 2009 to 2019. The results showed a higher co-agglomeration level significantly increased environmental pollution. The transmission and upgrading from secondary to tertiary industries exhibited a remarkable intermediary role, yet the credit scale formed a nonlinear threshold effect. Both industrial structure optimization and credit scale expansion contributed to environmental protection. Nevertheless, the path of “industrial co-agglomeration → technological progress → environmental protection” was not obvious, and the positive externalities of technology need to be strengthened. These findings provide viable insights for the implementation of financial and manufacturing integration and green development.
•Mediation mechanisms and threshold effects are performed.•Water resources are identified as a significant instrumental variable.•Financial and manufacturing co-agglomeration increase environmental pollution.•The partial mediating effect of industrial structure is verified.•The expansion of credit scales reduces environmental pollution.
was domesticated from
. The long-term domestication of the silkworm has brought about many remarkable changes to its body size and cocoon shell weight. However, the molecular mechanism underlying the ...improvement in the economic characteristics of this species during domestication remains unclear. In this study, we found that a transposable element (TE)-Bm1-was present in the upstream regulatory region of the
(Max-like protein X) gene in wild silkworms but not in all domesticated silkworms. The absence of Bm1 caused an increase in the promoter activity and mRNA content of
. Mlx and its partner Mondo belong to the bHLHZ transcription factors family and regulate nutrient metabolism. RNAi of
and
decreased the expression and promoter activity of glucose metabolism-related genes (
(
),
(
), and
(
)), lipogenic genes (
(
) and
(
)), and glutamine synthesis gene (
(
)). Furthermore, the transgenic overexpression of
and
in the fat body of silkworms increased the larval body size, cocoon shell weight, and egg number, but the silencing of the two genes resulted in the opposite phenotypes. Our results reveal the molecular mechanism of
selection during domestication and its successful use in the molecular breeding of
.
This study aimed to evaluate the effectiveness of double ABCX-based psychotherapy for psychological distress during
fertilization-embryo transfer (IVF-ET) among a female group (FG), couple group ...(CoG) and control group (CG).
A total of 201 women undergoing their first IVF-ET cycle were randomized into three groups. The 6-session intervention was delivered at each visit to the IVF clinic. The primary outcomes were depression and anxiety, and the secondary outcomes included sleep quality, serum cortisol (nmol/L) levels and the clinical pregnancy rate, which were assessed before and after the intervention.
The group-by-time effects were significant for depression, anxiety, sleep quality and serum cortisol levels, with larger effect sizes in the FG than in the CoG. There was no significant difference in the pregnancy rate among the three groups.
Psychotherapy effectively mitigated psychological distress, suggesting greater effectiveness for couples undergoing IVF couples than for women only. It is structured and easy to use during the IVF treatment cycle.
https://register.clinicaltrials.gov (NCT03931187, retrospectively registered on April 23, 2019).
Multi-population cohorts offer unprecedented opportunities for profiling disease risk in large samples, however, heterogeneous risk effects underlying complex traits across populations make ...integrative prediction challenging. In this study, we propose a novel Bayesian probability framework, the Prism Vote (PV), to construct risk predictions in heterogeneous genetic data. The PV views the trait of an individual as a composite risk from subpopulations, in which stratum-specific predictors can be formed in data of more homogeneous genetic structure. Since each individual is described by a composition of subpopulation memberships, the framework enables individualized risk characterization. Simulations demonstrated that the PV framework applied with alternative prediction methods significantly improved prediction accuracy in mixed and admixed populations. The advantage of PV enlarges as genetic heterogeneity and sample size increase. In two real genome-wide association data consists of multiple populations, we showed that the framework considerably enhanced prediction accuracy of the linear mixed model in five-group cross validations. The proposed method offers a new aspect to analyze individual's disease risk and improve accuracy for predicting complex traits in genotype data.
Social recommender systems (SRS) aim to study how social relations influence users’ choices and how to use them for better learning users embeddings. However, the diversity of social relationships, ...which is instructive to the propagation of social influence, has been rarely explored. In this paper, we propose a graph convolutional network based representation learning method, namely multi-perspective social recommendation (MPSR), to construct hierarchical user preferences and assign friends’ influences with different levels of trust from varying perspectives. We further utilize the attributes of items to partition and excavate users’ explicit preferences and employ complementary perspective modeling to learn implicit preferences of users. To measure the trust degree of friends from different perspectives, the statistical information of users’ historical behavior is utilized to construct multi-perspective social networks. Experimental results on two public datasets of Yelp and Ciao demonstrate that the MPSR significantly outperforms the state-of-the-art methods. Further detailed analysis verifies the importance of mining explicit characteristics of users and the necessity for diverse social relationships, which show the rationality and effectiveness of the proposed model. The source Python code will be available upon request.