Method and apparatus for asynchronous data fusion, storage medium and electronic device

An apparatus for asynchronous data fusion, comprising:

• a LiDAR data processing module, configured to obtain current frame LiDAR data, and determine current frame LiDAR three-dimensional embeddings of the current frame LiDAR data;
• a temporal fusion processing module, configured to determine a previous frame fused hidden state, and perform a temporal fusion process based on the previous frame fused hidden state and the current frame LiDAR three-dimensional embeddings to generate a current frame temporary hidden state and a current frame output result, wherein the previous frame fused hidden state is generated based on a previous frame temporary hidden state and previous frame camera three-dimensional embeddings;
• and
• a camera data processing module, configured to obtain current frame camera data, determine current frame camera three-dimensional embeddings of the current frame camera data, and generate a current frame fused hidden state based on the current frame camera three-dimensional embeddings and the current frame temporary hidden state, wherein the current frame fused hidden state is used for a process of performing a temporal fusion process based on the current frame fused hidden state and next frame LiDAR three-dimensional embeddings to generate a next frame temporary hidden state.

A computer storage medium storing computer executable instructions, wherein the computer program is used to, when being executed by a processor, perform:

• obtaining current frame LiDAR data, and determining current frame LiDAR three-dimensional embeddings of the current frame LiDAR data;
• determining a previous frame fused hidden state, and performing a temporal fusion process based on the previous frame fused hidden state and the current frame LiDAR three-dimensional embeddings to generate a current frame temporary hidden state and a current frame output result, wherein the previous frame fused hidden state is generated based on a previous frame temporary hidden state and previous frame camera three-dimensional embeddings;
• and
• obtaining current frame camera data, determining current frame camera three-dimensional embeddings of the current frame camera data, and generating a current frame fused hidden state based on the current frame camera three-dimensional embeddings and the current frame temporary hidden state, wherein the current frame fused hidden state is used for a process of performing a temporal fusion process based on the current frame fused hidden state and next frame LiDAR three-dimensional embeddings to generate a next frame temporary hidden state.