Freiburg Groceries Dataset
C++ Python Makefile
Switch branches/tags
Nothing to show
Fetching latest commit…
Cannot retrieve the latest commit at this time.

The Freiburg Groceries Dataset

The Freiburg Groceries Dataset consists of 5000 256x256 RGB images of 25 food classes. Examples for each class can be found below. The paper can be found here and the dataset here.

Example images:

Example images

Download the Dataset and Setup the Evaluation

First, clone the repository and navigate to the src directory:
[user@machine folder] git clone
[user@machine folder] cd freiburg_groceries_dataset/src

You can download the dataset with python3:
[user@machine src] python

Then, edit and specify the path to your caffe installation, the path to your cuda installation and the gpu that should be used for training.

To install the evluation software the following libraries are required: caffe, cuda, boost, python3, numpy. The evaluation software is partly written in C++. To clone the repo and build the evluation run
[user@machine src] python
This also downloads the bvlc_reference model we use for finetuning if necessary. Make sure you are in the src directory, as all paths are relative from there.


You can start training with
[user@machine folder] python
This creates the lmdbs, trains the 5 splits and evaluates them on the corresponding test set. This includes computing the accuracy for each class and producing a confusion matrix. It also links the misclassified images for each class and names them to contain the class they were confused with.

Some Examples for all Classes

Class overview images