China develops AI for missile targeting of stealth aircraft
Chinese researchers developed an AI system optimized for resource-constrained deployment in missile guidance systems to identify thermal signatures of stealth aircraft including U.S. F-22 and F-35 fighters. The model achieved over 90% accuracy on target aircraft in laboratory tests and is intended for integration into future air-to-air missile systems.
What happened
Researchers from Beijing Institute of Technology and China Academy of Air-Launched Missiles developed an AI model designed to recognize the infrared thermal signatures of stealth fighter aircraft including the F-22 and F-35.
The team trained the model on 3,245 infrared images captured by a scanning system installed in a missile, including three categories of airborne targets with simulated representations of the F-22 and F-35 as well as a loitering munition.
The model achieved 97.1% maximum accuracy in image classification during laboratory tests, with a rate near 90% specifically for the two fighter jet targets. The researchers compressed the model to 16.1% of parameters used by previous methods and reduced computational requirements to 19.2% of earlier approaches.
Reported impact
The AI is intended for integration into future air-to-air missile infrared guidance systems to enable more effective targeting of stealth aircraft that rely on thermal signature reduction as a primary evasion tactic.
Index analysis
Stealth aircraft such as the F-22 and F-35 are designed to reduce radar signature detection, but their thermal emissions from engines, exhaust gases, and airframe heating remain detectable by infrared sensors.
The challenge addressed by this AI is distinguishing actual aircraft thermal signatures from decoy flares and countermeasures that missiles use for evasion, which must be resolved in milliseconds during engagement.
Classification
Sources and evidence
This record was researched and written by the Index, based on the following original reporting.