Computes loss for delta output. Only relevant if rezoom layers are used.

Take the txt file and net configuration and create a generator that outputs a jittered version of a random image from the annolist that is mea

Set a value for the configuration. Parameters ---------- cfg_name : str env_name : str default : str cfg : function

Produce weights for a reduced fully-connected layer. FC8 of VGG produces 1000 classes. Most semantic segmentation task require much

Helper to create an initialized Variable with weight decay. Note that the Variable is initialized with a truncated normal distributio

Build the VGG model using loaded weights Parameters ---------- rgb: image batch tensor Image in rgb shap.

Analyze a trained model. This will load model files and weights found in logdir and run a basic analysis. Parameters ----------

Train model for a number of steps. This trains the model for at most hypes['solver']['max_steps']. It shows an update every utils.cfg.st

Load the model and run evaluation. Current Version runs the evaluation defined in network.evaluation and prints the output to std out.

force the Aspect ratio

Load dictionary which maps colors to classes as well as the default class. The classes are integers with values which range from 0 to N-1, w

initialize class names

Creating the training folder and copy all model files into it. The model will be executed from the training folder and all outputs will

======================================================================= FUNCTIONS TO LOAD DETECTION AND GROUND TRUTH DATA ONCE, SAVE RESULTS =========

Download the data if it isn't downloaded by now. Parameters ---------- hypes : dict Hyperparameters

Reads a kitti annotation file. Args: label_file: Path to file Returns: Lists of rectangels: Cars and don't care area.