This paper presents an overview on current research in feature recognition using artifi-cial neural network (ANN) techniques. ANN-based feature recognition approaches can eliminate some drawbacks of conventional feature recognition methods, and therefore have become an important research focus. The paper discusses the four main issues re-lated to ANN-based feature recognition techniques with the current systems, namely topology of neural network, input representations, training methods and output formats. Finally, it summarises limitations and future research directions.
|Number of pages||6|
|Publication status||Published - 2000|
|Event||16th National Conference on Manufacturing Research - London|
Duration: 5 Sep 2000 → 7 Sep 2000
|Conference||16th National Conference on Manufacturing Research|
|Period||5/09/00 → 7/09/00|