I'm working on object detection using dlib, I was going through python implementation. I tested couple of examples from dlib python examples, specially i worked on train_object_detector.py. this works well. Now I would like to train same data model on CNN based object detector mode But i could not find python implementation for training CNN using python, but there is c++ example. (dnn_mmod_ex.cpp). I think it seems i'm missing something or python implementation is not available?
if python implementation is not available, then should i switch to c++ for CNN based object detector training
Yes, use C++ for CNN training. The dlib DNN tooling is meant to be used from C++ and so uses C++11 features that can't be represented in Python.
Related
There are ~500k code snippets written in proprietary language that I have to port to a new system also using its own proprietary language. I have the following with me
Vocabulary and grammar of source and destination languages
Sample of 1500 converted rules (for training if required) of different complexity
I am not looking for 100% automation but may be a transpiler that may automate part of it. Can it be done using NLP? Have already gone through this, this, Rascal , Haxe and Spoofax. I could not find much documentation on how to create a custom tranpiler.
Any help is appreciated. Thank you!
I'm looking into performing object detection (not just classification) using CNNs; I currently only have access to Windows platforms but can install Linux distributions if necessary. I would like to assess a number of existing techniques, but most available code is for Linux.
I am aware of the following:
Faster RCNN (CNTK, Caffe w/ Matlab)
R-CNN (Matlab with loads of
toolboxes)
R-FCN (Caffe w/ Matlab)
From what I can see, there are no TensorFlow implementations for Windows currently available. Am I missing anything, or do I just need to install Ubuntu if I want to try more?
EDIT: A windows version of YOLO can be found here: https://github.com/AlexeyAB/darknet
There is a Tensorflow implementation for Windows but honestly it's always 1-2 steps behind Linux.
One thing I would add to your list (which was not already mentioned) is MaskRCNN
You can also look for some implementation of it for Tensorflow like this one: https://github.com/CharlesShang/FastMaskRCNN
Since the documentation of OpenCV 3.0 machine learning library is obsolete (there is also a bug in svm->load), I'd like to use OpenCV 2.4.9 machine learning library instead. However, my entire program in which I want to use machine learning algorithms is written based on OpenCV 3.0 and it is a long code.
Because of conflicts, I can't link to two different OpenCV libraries in the same project. One idea is to define another project in the same solution in which I can link to OpenCV 2.4.9 but I don't know how to do this properly (apparently, this second project can't be a static library because it's difficult to link a static library to another library). I should be able to pass values between the two projects (i.e. between my original program using OpenCV 3.0 and machine learning algorithms using OpenCV 2.4.9).
I was looking into Tkinter, but I was wondering if there are any other packages that help with GUI. Even though Java is more robust than python, python is faster to write and would be very helpful if I am just trying to create a quick application. I usually use .Swing and .awt in java to make all of my applications.
Also, Is there something similar to JWS for python? I am looking to add a link on a website that will fire up a python coded application
Thanks
(P.S. I'm using Python 3.2.2)
I love using tkinter with python. It is pretty easy to do, there are lots of examples, and numerous questions with answers on here too! Also, 90% of people with python have tkinter already on their system, making distribution pretty straight forward. I also have had success using py2exe to create windows executables.
Here is a gui that I wrote in all of ~10 minutes. Once you understand the syntax it is very easy.
And another more elaborate one that I wrote with vtk and matplotlib embedded into the gui for 3D and 2D graphics! The question I asked with this.
Yes, python and tkinter are excellent for developing GUIs. While this is true of most scripting languages, I think these two make a particularly good combination.
For most GUIs, functionality is more important than form (ie: eye candy is not top priority). Even then, it's possible to make very nice looking GUIs with Tkinter. Tk has been my go-to toolkit for close to twenty years. It is most easily used with Tcl since that is its native language, but there's no denying the advantage python has over Tcl as far as built-in functionality goes.
That being said, all of the toolkits available to a python programmer are good. Tkinter's main advantages are that it is extremely easy to use and you probably already have it installed, so the barrier for entry is very low.
I started learning Standard ML recently out of curiosity. So what I know is that is has an efficient compiler (MLton) which allows us to freely use abstractions without worrying about performance.
It would be perfect if I could do some GUI programming with Standard ML, too. Is there anything like Gtk, Qt, or WxWidgets binding for Standard ML?
Thank you very much.
Here's a paper describing the mGTK project, which is a Standard ML binding to GTK. There are also a couple of interfaces to OpenGL, although this is more 2 and 3D graphics-oriented rather than specifically a GUI toolkit. eXene is a toolkit for X Windows. sml_tk is an interface to Tk.
There are bindings for foreign toolkits, but the eXene toolkit was designed to be native for SML and to exploit the features of Concurrent ML. I used it years ago and found it a very smooth fit for the language and a pleasure to use. But it does not have the ginormous library of components that you find in the more widely used toolkits.
Giraffe Library provides SML bindings to libraries that support GObject Introspection, which includes GTK. At the time of writing, bindings are available for the GTK 3 stack. Although the cairo graphics library does not support GObject Introspection, Giraffe Library provides bindings to some cairo functions.
Although SML has a very precise definition, the definition does not include an FFI, and there is considerable variation between compilers. At the time of writing, Giraffe Library supports only the FFIs of MLton and Poly/ML.