This is an implementation of the Differential Edit Distance (DED) metric, a unidimensional measure that has been developed to evaluate video scenes segmentation results. It compares the scene segmentation results of an automatic scene segmentation system with the ground-truth scene segmentation and estimates their similarity in terms of the number of shots that need to change scene in order to make the two segmentations identical. The software input is two .txt files containing the segmentation under comparison and the output is their distance, which can be also saved or appended in a .txt file. The software can be also used to compare any label strings that correspond to sets of well-ordered objects.
Windows 32-bit version
P. Sidiropoulos, V. Mezaris, I. Kompatsiaris, J. Kittler, "Differential Edit Distance: A metric for scene segmentation evaluation"
IEEE Transactions on Circuits and Systems for Video Technology, accepted for publication.
THIS SOFTWARE IS PROVIDED BY THE AUTHOR "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
This work was supported by the EU FP7 projects GLOCAL (FP7-248984).
Feedback is kindly appreciated.