In many applications, real-time image analysis is mandatory. An industry-robot, which needs five seconds to recognize objects on an image is mostly very limited in usage. On the other hand latest applications require detecting objects and positons as accurate as possible. Using some high-performance GPUs optimized for deep learning and latest research advances like YOLO, […]
Especially in the field of Machine Learning, you will read about an approach called “Data Augmentation”. This idea is great to boost the performance of your Machine Learning Models. First at all, we want to clarify what Data Augmentation means. Data Augmentation is an approach to generate more training samples for your model without actually […]
Today we have a look at how you can create a text recognition software with simple methods without being dependent on Tesseract and similar. We will get to know the following: Semantic Local Binary Patterns (S-LBP): As feature extraction method1 Histograms: As method to efficiently merge/collect features k-Nearest-Neighbor Classifier: As simple method to classify samples […]
This post summarizes the ideas behind pattern recognition and explains a few words that are very often used in pattern recognition. Specifically, we get to know the following: What is pattern recognition and where is it used? Machine Learning: How can computers learn? Trainings set and test set: Why we need two different data sets […]
So this is the final day. There was not much I did today. I only cleaned up my code and did some minor fixes/improvements. I am happy that the first part of my project kind of succeeded. Of course there are some problems left (e.g. grey/white traffic signs are only found in around 50%), but […]
YAY! I kind of did it! I created a reasonably fast traffic sign/light detector. The software truely does not work perfectly and also detection of gray/white signs is still quite hard. However it basically works, especially for red, blue and yellow traffic signs using only really limited hardware. There is still much room left to […]
The problem regarding a faster region proposal seems to be solved. The SVM + EA combination did not work fast enough. So I decided quite early to try solving my problem on an other way. I took a look into a bunch of quite old papers, where people tried to detect and recognize traffic signs […]
So I finally found an architecture which seem to work pretty good. We also continued annotating images taken and labeled about 1000 more traffic signs/traffic lights. In total we now got about 3000 traffic signs/lights in our own database + 800 from the The German Traffic Sign Detection Benchmark. We applied a lot of data […]
Day #4 was kind of boring. I tried different NN architectures and kind of watched them training the whole day. I trained approximately 5 different architectures and the last architecture I tried seems to work the best. Howeve even this NN does not work as good as I want it to work, so tomorrow the […]
Today I was able to test my new CNN architecture and the extended dataset. However preprocessing and augmenting the annotated images took much longer than I expected it would (It took around of 1h manual postprocessing and ~3h automatic post-/preprocessing). I started training my new NN architecture. My goal was 120.000 iterations. This takes about […]