Abstract: A small-size vehicle detection deep learning model based on feature fusion of multi-scale modules is provided, which solves the problem of small-size vehicle image detection. The model includes a Backbone network, a Neck layer and a Head network, wherein a C2f_DCNv3 module based on the combination of deformable convolution v3 (DCNv3) and a cross stage feature fusion (C2f) module and an SPPF_LSKA module based on the combination of a spatial pyramid pooling fast (SPPF) layer and a large separable kernel attention (LSKA) module are introduced into the Backbone network; a C2f_SCConv module based on the combination of spatial and channel reconstruction convolution (SCConv) and a C2f module is introduced into the Neck layer; and a multi-scale kernel detection (MSK_Detect) module is introduced into the Head network.

Abstract: The invention relates to the technical field of welding machine protection, specifically a protective device for interlocking welding machines, a design method, and a shield. The device includes: an inductive detection module to detect the relative position between the shield and a human head; a position detection module to detect the filter lens position relative to the shield; a determination module connected to both detection modules to check if all parameters are within a preset range, converting them into an unlocking signal if so; and a transmission module to send this signal to the welding machine, allowing operator control after unlocking. The invention enhances safety, reduces injury risks, and offers reliable protection, with advantages like high safety, automatic control, and effective interlocking.

Abstract: The present disclosure provides a remote host monitoring method based on chip-level privacy-preserving computation (PPC), including: S1: allowing monitoring software in a user-side host to start in a chip-level trusted execution environment (TEE); S2: determining whether the user-side host locally stores valid private data, directly going to step S4 if yes, or otherwise, going to step S3; S3: establishing a secure connection with a supervisor and capturing private data; and S4: allowing monitoring software running in the TEE to execute a related monitoring instruction based on the private data, encrypting and signing a monitoring result, and transmitting the monitoring result to the supervisor. The present disclosure ensures validity, tamper resistance and security of monitoring information of the user with a TEE based on PCC, encrypts and signs the monitoring information based on an encryption key and a signature key of the supervisor, can locally store the monitoring information.

Abstract: The present disclosure provides a remote host monitoring method based on chip-level privacy-preserving computation (PPC), including: S1: allowing monitoring software in a user-side host to start in a chip-level trusted execution environment (TEE); S2: determining whether the user-side host locally stores valid private data, directly going to step S4 if yes, or otherwise, going to step S3; S3: establishing a secure connection with a supervisor and capturing private data; and S4: allowing monitoring software running in the TEE to execute a related monitoring instruction based on the private data, encrypting and signing a monitoring result, and transmitting the monitoring result to the supervisor. The present disclosure ensures validity, tamper resistance and security of monitoring information of the user with a TEE based on PCC, encrypts and signs the monitoring information based on an encryption key and a signature key of the supervisor, can locally store the monitoring information.