Sound source localization for robots - Revision history

SimonePradolini at 11:32, 10 December 2015

2015-12-10T11:32:50Z

SimonePradolini at 11:30, 10 December 2015

2015-12-10T11:30:32Z

SimonePradolini: /* Existing systems */

2015-12-10T11:24:57Z

‎Existing systems

SimonePradolini: /* */

2015-12-10T11:21:39Z

‎

SimonePradolini: /* Pros and cons of existing approaches */

2015-12-10T11:11:13Z

‎Pros and cons of existing approaches

SimonePradolini at 10:38, 10 December 2015

2015-12-10T10:38:55Z

SimonePradolini: /* Existing systems */

2015-12-10T10:33:55Z

‎Existing systems

SimonePradolini: /* State-of-the-art */

2015-12-10T10:12:48Z

‎State-of-the-art

SimonePradolini: /* State-of-the-art */

2015-12-10T10:11:09Z

‎State-of-the-art

SimonePradolini at 09:44, 10 December 2015

2015-12-10T09:44:54Z

@@ Line 1: / Line 1: @@
-{{Project template
+{{Project
 |title=Sound source localization for robots
 |description=Analysis of the state-of-the-art of binaural systems for audio detection and localization in robotics
@@ Line 5: / Line 5: @@
 |students=SimonePradolini
 |status=Active
 |start=2015/10/08
-|number=1
+|type=Thesis
-|cfu=20
 |image=TESI.png
 }}

@@ Line 3: / Line 3: @@
 |description=Analysis of the state-of-the-art of binaural systems for audio detection and localization in robotics
 |tutor=MatteoMatteucci
 |start=2015/10/08
 |number=1

@@ Line 26: / Line 26: @@
 ===Existing systems===
 Here we will describes in few words the state-of-the-art
 ====Nakashima, Hiromichi, and Toshiharu Mukai. "3D sound source localization system based on learning of binaural hearing." [http://ieeexplore.ieee.org/xpls/abs_all.jsp?arnumber=1571695]====

@@ Line 33: / Line 33: @@
 In this work, Deleforge uses implicitly the HRTF jointly with ITD and IPD. The sound incoming is acquired using an active binaural robot, a system with a real human head shape that can move in horizontal and vertical planes (pan and tilt movements). The sound source is in a fixed point, instead the robot change its orientation, moving in pan and tilt. The utilization of a humanoid head make implicit that they are filtering the input signal via the HRTF of this robot. The main idea of the 2D separation and localization algorithm is  a probabilistic model, a Gaussian Mixture model, solved via Expectation Maximization using as cues ITD and IPD.
-==== ====
+====Willert, Volker, et al. "A probabilistic model for binaural sound localization."[http://ieeexplore.ieee.org/xpls/abs_all.jsp?arnumber=1703643] ====
 ==Pros and cons of existing approaches==

@@ Line 36: / Line 36: @@
 ==Pros and cons of existing approaches==
+We decided to focus our attention on two researches: the results obtained by the Perception group of the INRIA [https://team.inria.fr/perception/] and by the Institute of Automatic Control of the Darmstadt University of Technology [http://www.rmr.tu-darmstadt.de/rmr/aktuelles_rmr/index.en.jsp]. We would like to analyze their work and make some new proposal to achieve better performances. Then we can try to use this knowledge to create our general purpose sound source localization algorithm. Our goal is to create a system that can work on any robot in any environment.
-−
 ==Essential bibliography==

← Older revision		Revision as of 10:38, 10 December 2015
Line 34:		Line 34:

	==== ====		==== ====
		+
		+	==Pros and cons of existing approaches==
		+
		+

	==Essential bibliography==		==Essential bibliography==

← Older revision		Revision as of 10:33, 10 December 2015
Line 25:		Line 25:

	===Existing systems===		===Existing systems===
		+	Here we will describes in few words the state-of-the-art
		+
		+	====Nakashima, Hiromichi, and Toshiharu Mukai. "3D sound source localization system based on learning of binaural hearing." [http://ieeexplore.ieee.org/xpls/abs_all.jsp?arnumber=1571695]====
		+	In this paper, Nakashima proposes a robot with two acquiring sensors (two microphones) and with 2 humanoid pinnas to achieve both horizontal and vertical detection and localization. He uses as features ITD for the azimuth and spectral features for the elevation. In this way, Nakashima creates a physical model, a sort of simplified version of an HRTF creating a transfer function between the incoming signal and the "humaniod" robot sensors composed only by ears and a quasi-spherical main corpus. The training set is done mapping the minimum of this mixed feature vector (composed by ITD and spectral content) feeding the system with white noise emitted from every azimuth and elevation position through a neural system.
		+
		+	====Deleforge, Antoine, and Radu Horaud. "The cocktail party robot: Sound source separation and localisation with an active binaural head."[http://dl.acm.org/citation.cfm?id=2157834]====
		+	In this work, Deleforge uses implicitly the HRTF jointly with ITD and IPD. The sound incoming is acquired using an active binaural robot, a system with a real human head shape that can move in horizontal and vertical planes (pan and tilt movements). The sound source is in a fixed point, instead the robot change its orientation, moving in pan and tilt. The utilization of a humanoid head make implicit that they are filtering the input signal via the HRTF of this robot. The main idea of the 2D separation and localization algorithm is a probabilistic model, a Gaussian Mixture model, solved via Expectation Maximization using as cues ITD and IPD.
		+
		+	==== ====

	==Essential bibliography==		==Essential bibliography==

← Older revision		Revision as of 10:12, 10 December 2015
Line 23:		Line 23:

	For our primary task, that is a simpler and stable source localization algorithm for robotics with no willing of emulating human perception system, working with ILD/ITD could be sufficient. If we would like to go in deepth with the analysis of human physiology, human perception or environment analysis, HRTF could be a better choice in future.		For our primary task, that is a simpler and stable source localization algorithm for robotics with no willing of emulating human perception system, working with ILD/ITD could be sufficient. If we would like to go in deepth with the analysis of human physiology, human perception or environment analysis, HRTF could be a better choice in future.
		+
		+	===Existing systems===

	==Essential bibliography==		==Essential bibliography==

@@ Line 11: / Line 11: @@
 ==State-of-the-art==
 Robotics audition and active perception of sound source are nowadays a hard task. The common starting point of different existing algorithm is the knowledge of some psychoacoustics measurement. These main features are ITD (Interaural Time Difference) and ILD (Interaural Level Difference), and they describe how humans can localize a sound source.
@@ Line 16: / Line 17: @@
 ILD measures the difference in pressure level (dB) between the two ears. As for ITD, ILD depends on the frequency content of the sound wave, on the incident angle and on physical structure of the human perceiving. For these reasons, the most useful IPD measurements are at high frequency (many state-of-the-art system use the same threshold as ITD, 1500Hz).
 ==Essential bibliography==

@@ Line 17: / Line 17: @@
 ILD measures the difference in pressure level (dB) between the two ears. As for ITD, ILD depends on the frequency content of the sound wave, on the incident angle and on physical structure of the human perceiving. For these reasons, the most useful IPD measurements are at high frequency (many state-of-the-art system use the same threshold as ITD, 1500Hz).
+==Essential bibliography==
-----
-−
-−
-Essential bibliography:
-−
 Haykin, Simon, and Zhe Chen. "The cocktail party problem." Neural computation 17.9 (2005): 1875-1902 [http://www.mitpressjournals.org/doi/abs/10.1162/0899766054322964#.Vi5BNux_Oko]