0

As far as I know, because Python is an interpreted language, the same code runs considerably faster in c#, c++ or c comparing to Python.

Questions:

Would a YOLO adaptation to c# be faster in a real-time situation? (like https://github.com/mentalstack/yolov5-net)

How does the interpreter actually impact YOLOv5 and YOLOv1-4?

Caio
  • 1
  • Welcome to Stack Overflow. Please read [ask] and https://stackoverflow.com/help/on-topic and note well that this is **not a discussion forum**. We expect a *specific* question. It's impossible to answer "why is X the fastest?", because a) you do not say what *task* it is supposedly "fastest" at; b) you do not say *what it is being compared to (fastest *out of what options*?). Note that we expect **one** question at a time, and we need them to be properly focused - "how does the interpreter impact ?" is not feasible to explain. – Karl Knechtel Jun 09 '22 at 22:02
  • As an aside, "compiled vs interpreted languages" is not a real distinction that has made sense for many years. Python (in the reference implementation) is compiled to an interpreted bytecode form - the exact same way that Java and .NET (including C#) work normally. Also, in each of those cases it *isn't a property of the language*; it's a property of the software. There is nothing preventing anyone from trying to "compile" Python, and on the flip side, [interpreters for C exist](https://stackoverflow.com/questions/584714/is-there-an-interpreter-for-c). – Karl Knechtel Jun 09 '22 at 22:03
  • Well, Java and C# are both compiled to a bytecode, but that bytecode is JIT compiled to machine language on startup. That's quite different from CPython, which interprets the byte code. However, it is wildly incorrect to say "the same code runs considerably faster" in C or C++. The difference is rarely very large. – Tim Roberts Jun 09 '22 at 22:53

0 Answers0