Research Keyword: Neural networks

Fully integrated hybrid multimode-multiwavelength photonic processor with picosecond latency

Researchers created a tiny optical computer chip that processes data much faster than electronic computers. The chip uses different colors and patterns of light to handle multiple signals at once, mimicking how fiber optic cables can send information through different pathways simultaneously. It successfully unscrambled mixed-up signals in real-time and removed jamming interference from communication signals, all with a processing time of just 30 picoseconds—trillions of times faster than blinking.

Read More »

Application of ATR-FTIR and FT-NIR spectroscopy coupled with chemometrics for species identification and quality prediction of boletes

Researchers developed a fast and non-destructive method to identify different types of edible boletes and assess their nutritional quality by analyzing their amino acid content. Using special spectroscopy techniques combined with computer analysis, they achieved perfect accuracy in identifying five bolete species and could predict the amino acid content that contributes to flavor and nutrition. This breakthrough provides consumers with better protection against accidentally purchasing toxic mushroom species that look similar to edible ones, while helping food producers quickly assess quality without lengthy lab testing.

Read More »
Scroll to Top