Developmental Network Two, Its Optimality, and Emergent Turing Machines

    公开(公告)号:US20190392321A1

    公开(公告)日:2019-12-26

    申请号:US16265212

    申请日:2019-02-01

    IPC分类号: G06N3/08

    摘要: This invention includes a new type of neural network that is able to automatically and incrementally generate an internal hierarchy without a need to handcraft a static hierarchy of network areas and a static number of levels and the static number of neurons in each network area or level. This capability is achieved by enabling each neuron to have its own dynamic inhibitory zone using neuron-specific inhibitory connections.

    Starch-oil sizing for glass fibers
    4.
    发明授权
    Starch-oil sizing for glass fibers 失效
    用于玻璃纤维的淀粉 - 油上浆

    公开(公告)号:US5393335A

    公开(公告)日:1995-02-28

    申请号:US52323

    申请日:1993-04-23

    摘要: An aqueous starch-oil sizing composition is provided which produces improved processibility in woven and non-woven applications. The sizing has low viscosity starch, a lubricant which is a mixture of oil and wax where the wax is present in an amount of at least twice the oil, cationic lubricants, and a humectant. The size also includes an organo-functional silane coupling agent such as gamma-glycidoxypropyltrimethoxy silane which may be hydrolyzed. Emulsifiers, defoamers, and biocides may be present. Strands of glass fiber which have been treated with this size have generally shown a reduced tendency to shed the size from the strands, a reduction in the buildup of broken filaments or "fuzz" on processing equipment and reduced strand breakage.

    摘要翻译: 提供了一种水性淀粉 - 油上胶组合物,其在织造和非织造应用中产生改进的加工性能。 上浆具有低粘度淀粉,润滑剂是油和蜡的混合物,其中蜡的存在量至少为油的两倍,阳离子润滑剂和湿润剂。 该尺寸还包括可水解的有机官能的硅烷偶联剂,例如γ-环氧丙氧基丙基三甲氧基硅烷。 可能存在乳化剂,消泡剂和杀生物剂。 已经用这种尺寸处理的玻璃纤维束通常显示出从股线上减小尺寸的趋势,减少在加工设备上破裂的丝或“绒毛”的积聚和减少的线断裂。

    IMAGING AND FORMING METHOD USING PROJECTION OPERATION AND BACK PROJECTION METHOD

    公开(公告)号:US20180153205A1

    公开(公告)日:2018-06-07

    申请号:US15570749

    申请日:2016-04-23

    申请人: Xiang Wu

    发明人: Xiang Wu

    IPC分类号: A23P10/00 B33Y50/00 A23L15/00

    摘要: A method of the present invention comprises: manufacturing a two-dimensional or three-dimensional real image using a projection operation and back projection method, and also comprises: completing two-dimensional display, three-dimensional display, two-dimensional printing and three-dimensional printing using the real image. The present invention belongs to the field of flat-panel display, the field of 3D stereographic display, the technical field of printing, the field of rapid prototyping, the field of additive manufacturing, and the field of 3D printing. The method is similar to a projection data collection and back projection reconstruction method in the computed tomography (CT) technology. The achievement of the computed tomography (CT) technology is the projection data collection of real objects and reconstruction of digitized tomographic images, so as to convert the real objects into virtual data. Projection data collection is replaced with projection operation, the digitized back projection reconstruction method is replaced with the real back projection method, so as to convert virtual data into real objects or real images. The adopted projection rays comprise light, electromagnetic waves, high-energy rays, particle flows, sound waves, shock waves, currents or chemical waves.

    Net chain driving structure
    6.
    发明授权

    公开(公告)号:US11618630B2

    公开(公告)日:2023-04-04

    申请号:US17497986

    申请日:2021-10-11

    申请人: Xiang Wu

    发明人: Xiang Wu

    摘要: A net chain driving structure includes a conveying net chain. The conveying net chain includes a plurality of chain links successively inserted in a conveying direction, a plurality of lugs spaced apart from each other are arranged on two sides of the chain links, an end portion of the lug is an arcuate face, a meshing groove is disposed on the chain link, the meshing groove is spliced by an open slot provided on a back side of the chain link and the lugs on the adjacent chain links, a gear in transmission meshing with the meshing groove is disposed on each end of the conveying net chain, a tangent plane tangential to the arcuate face is disposed on the arcuate face of the lug, and transmission contact toothed surfaces of the tangent plane and the gear are attached to each other in parallel during meshing transmission.

    Robot that Concurrently Learns Recognition and Synthesis while Developing a Motor

    公开(公告)号:US20230034287A1

    公开(公告)日:2023-02-02

    申请号:US17379344

    申请日:2021-07-19

    申请人: Juyang Weng Xiang Wu

    发明人: Juyang Weng Xiang Wu

    IPC分类号: G06N3/00 G06N3/04 G06N3/08

    摘要: Traditionally, learning speech synthesis and speech recognition were investigated as two separate tasks. This separation hinders incremental development for concurrent synthesis and recognition, where partially-learned synthesis and partially-learned recognition must help each other throughout lifelong learning. This invention is a paradigm shift—we treat synthesis and recognition as two intertwined aspects of a lifelong learning robot. Furthermore, in contrast to existing recognition or synthesis systems, babies do not need their mothers to directly supervise their vocal tracts at every moment during the learning. We argue that self-generated non-symbolic states/actions at fine-grained time level help such a learner as necessary temporal contexts. Here, we approach a new and challenging problem—how to enable an autonomous learning system to develop an artificial motor for generating temporally-dense (e.g., frame-wise) actions on the fly without human handcrafting a set of symbolic states. Here the artificial motor corresponds to a combination of a multiplicity of robotic effectors, including, but not limited to, speaking, singing, dancing, riding a bike, swimming, and driving a car. The self-generated states/actions are Muscles-like, High-dimensional, Temporally-dense and Globally-smooth (MHTG), so that these states/actions are directly attended for concurrent synthesis and recognition for each time frame. Human teachers are relieved from supervising learner's motor ends. The Candid Covariance-free Incremental (CCI) Principal Component Analysis (PCA) is applied to develop such an artificial speaking motor where PCA features drive the motor. Since each life must develop normally, each Developmental Network-2 (DN-2) reaches the same network (maximum likelihood, ML) regardless of randomly initialized weights, where ML is not just for a function approximator but rather an emergent Turing Machine. The machine-synthesized sounds are evaluated by both the neural network and humans with recognition experiments. Our experimental results showed learning-to-synthesize and learning-to-recognize-through-synthesis for phonemes. This invention corresponds to a key step toward our goal to close a great gap toward fully autonomous machine learning directly from the physical world.

    Privacy Protection Query Language PQL and System Thereof

    公开(公告)号:US20220100887A1

    公开(公告)日:2022-03-31

    申请号:US17143182

    申请日:2021-01-07

    IPC分类号: G06F21/62 G06F16/2455

    摘要: The invention discloses privacy protection query language PQL and system thereof, comprising PQL statement and system, the system comprises parsing module, query module and noise-injection module; the parsing module comprises lexical analyzer and syntactic analyzer; user inputs PQL statement according to predetermined semantic and syntactic rules, after parsing module receives the PQL statement, lexical analyzer performs error checks on the PQL statement, correct results are sent to syntactic analyzer; the syntactic analyzer performs grammatical and semantic checks on PQL statement, generates a mapping table and a parameter table with correct results; after receiving mapping table, the query module encapsulates the same into SQL statement, and verifies the encapsulated SQL statement; the noise-injection module obtains final query results of the query module, calculates noise injection sensitivity according to parameter table; the invention ensures the data privacy security and satisfies that the injected noise does not affect data availability.