Patents
Literature
Patsnap Eureka AI that helps you search prior art, draft patents, and assess FTO risks, powered by patent and scientific literature data.

4results about How to "Maintain ability" patented technology

Single-person oxygen generator

The single-person oxygen generator comprises a compressor for compressing air and an adsorption assembly for separating nitrogen and oxygen in the air, the adsorption assembly comprises two adsorption tanks, a circulation tank is arranged between the two adsorption tanks, and a spiral sleeve, a rotating rod, a feeding plate and a collecting tank are arranged in the circulation tank; the rotating rod is coaxially and slidably connected with the spiral sleeve, the feeding plate is in threaded connection with the spiral sleeve, the collecting tank is located below the feeding plate, and the feeding plate can move in the circulation tank to screen out broken bodies and powder in molecular sieve particles and change positions of the screened-out molecular sieve particles. The circulation tank of the oxygen generator is used for screening broken bodies and powder of molecular sieve particles, meanwhile, the feeding plate can move the screened molecular sieve particles from the air inlet end to the air outlet end, the adsorption energy efficiency of the molecular sieve at the air inlet end is kept, and the collecting tank in the circulation tank can collect and timely treat the screened broken bodies and powder.
Owner:SHENZHEN JIAGONG TECH CO LTD

Chimeric broad-spectrum oncolytic adenovirus with multiple mechanisms synergizing with and enhancing efficacy of immunotherapy, and application thereof in tumor treatment

PendingEP4349975A4High infection efficiencyreducing adhesion and uptakePeptide/protein ingredientsMicroorganism based processes
Provided are a chimeric oncolytic adenovirus simultaneously expressing IL-12, IFN-γ, and CCL5, and an application thereof in tumor treatment. A capsid protein hexon of the oncolytic adenovirus is formed from chimerizing hexon sequences of the two serotype viruses Ad5 and Ad48, and a fiber protein is formed from chimerizing fiber sequences of the two serotype viruses Ad5 and Ad11. The chimeric oncolytic adenovirus can activate intrinsic anti-cancer activity of a variety of viral structural proteins, the ability to infect tumor cells is increased while also ensuring that the virus itself avoids interception from pre-existing neutralizing antibodies and adhesion and uptake of hepatocytes, and an effect of killing cancer cells is enhanced.
Owner:VONCOLYTIC THERAPEUTICS CO LTD

Punching plug structure

The invention discloses a piercing plug structure which comprises a plug body, a plurality of hole channels are formed in the plug body, the hole channels are distributed in the outer wall of the plug body in a honeycomb shape, and filler is arranged in the hole channels so that heat conduction can be blocked or the hole walls of the hole channels can be supported through the filler; the filler at least comprises one or more of oxides, carbides, nitrides, halides, sulfides or high-temperature-resistant metals. The problems that an existing top is prone to deformation in a continuous high-temperature state, the heat insulation effect is poor, and the service life is short are solved.
Owner:CHANGCHUN LAIMAN OPTOELECTRONIC INSTRUMENT CO LTD

A method for alleviating double forgetting of visual language model and an evaluation method

PendingCN122174881Amaintain abilitySolving double forgetfulnessInference methodsNeural learning methodsLinguistic modelContinual learning
This invention proposes a continuous learning method and evaluation method to mitigate the double forgetting of visual language models. The method includes: adding two parallel paths to the backbone network of a pre-trained visual language model—one a task-agnostic expert path and the other a task-related expert group path—to obtain a model based on heterogeneous expert hybridization; training the model using a two-stage training strategy, including: a first stage, freezing the task-related expert group path and training only the task-agnostic expert path, using contrastive learning to solidify the pre-trained general knowledge of the visual language model; a second stage, freezing the task-agnostic expert path and training only the task-related expert group path, using a cross-entropy loss function to learn task-specific knowledge; and an inference stage, dynamically fusing the outputs of the two parallel paths through a hierarchical routing mechanism to achieve collaborative learning of pre-trained knowledge and task-specific knowledge.
Owner:NANJING UNIV +1