2005:Symbolic Melodic Similarity Results

Introduction

These are the results for the 2005 running of the Audio Melody Extraction task set.

Goal

To retrieve the most similar incipits from the RISM A/II collection, given one of the incipits as a query.

Dataset

558 MusicXML files of the incipits from the RISM collection. 11 incipits of this collection are chosen as queries. These MusicXML files were converted to MIDI files.

Result

The scores reported here are averaged over the 11 queries.

Rank	Participant	Average Dynamic Recall	Normalized Recall at Group Boundaries	Average precision (non-interpolated)	Precision at N documents (N is number of relevant document)	Input Data Format	Runtime (s)	Machine	Detailed Evaluation Result
1	Grachten, Arcos & Mántaras	65.98%	55.24%	51.72%	44.33%	MIDI	80.174 *	B0	GAM_eval.txt
2	Orio, N.	64.96%	53.35%	42.96%	39.86%	XML	24.610	B4	O_eval.txt
3	Suyoto & Uitdenbogerd	64.18%	51.79%	40.42%	41.72%	MIDI	48.133	B3	SU_eval.txt
4	Typke, Wiering & Veltkamp	57.09%	48.17%	35.64%	33.46%	MIDI	51240	B4	TWV_eval.txt
5	Lemström, Mikkilä, Mäkinen & Ukkonen (P3)	55.82%	46.56%	41.40%	39.18%	MIDI	10.007 *	B0	LMMU_P3_eval.txt
6	Lemström, Mikkilä, Mäkinen & Ukkonen (DP)	54.27%	47.26%	39.91%	36.20%	MIDI	10.106 *	B0	LMMU_DP_eval.txt
7	Frieler & Müllensiefen	51.81%	45.10%	33.93%	33.71%	MIDI	54.593	B4	FM_eval.txt