Rate versus synchrony code for human action recognition
نویسندگان
چکیده
We propose a bio-inspired feedforward spiking network modeling two brain areas dedicated to motion (V1 and MT), and we show how the spiking output can be exploited in a computer vision application: action recognition. In order to analyze spike trains, we consider two characteristics of the neural code: mean firing rate of each neuron and synchrony between neurons. Interestingly, we show that they carry some relevant information for the action recognition application. We compare our results to Jhuang et al. (2007) on the Weizmann database. As a conclusion, we are convinced that spiking networks represent a powerful alternative framework for real vision applications that will benefit from recent advances in computational neuroscience. Key-words: Spiking networks, bio-inspired model, motion analysis, V1, MT, human action recognition ∗ [email protected] † [email protected] ‡ [email protected] § [email protected] in ria -0 03 26 58 8, v er si on 1 3 O ct 2 00 8 Taux de décharge ou synchronisation pour la reconnaissance de mouvements humains Résumé : Nous proposons un réseau de neurones impulsionnels bio-inspiré qui modélise deux aires corticales dédiées au mouvement (V1 et MT), et nous montrons comment la sortie impulsionnelle peut être exploitée dans une application de vision par ordinateur: la reconnaissance d’action. Pour analyser les trains d’impulsions, nous considérons deux caractéristiques du code neural: le taux de décharge de chaque neurone et la synchronie entre les neurones. Nous montrons que chacun des deux codes véhicule une information pertinente pouvant être utilisée pour la reconnaissance. Pour cela, nous comparons nos résultats avec Jhuang et al. (2007) sur la base de séquence d’images Weizmann. En conclusion, nous sommes convaincus que les réseaux de neurones impulsionnels représentent une méthodologie alternative efficace pour les application de vision artificielle, qui bénéficiera des avancées récentes dans le domaine des neurosciences computationelles. Mots-clés : Réseaux de neurones à impulsions, modèles bio-inspirés, analyse de mouvement, V1, MT, reconnaissance de mouvement humain in ria -0 03 26 58 8, v er si on 1 3 O ct 2 00 8 Rate versus synchrony code for human action recognition 3
منابع مشابه
Can spike-based speech recognition systems outperform conventional approaches?
The field of automatic speech recognition (ASR) has advanced far enough in the past decade to produce numerous commercial applications such as the speech-driven telephone customer service menus now deployed by many companies. Unfortunately, these and other state-of-the-art ASR systems still pale in comparison to human performance, particularly in the presence of noise. Researchers have long bee...
متن کاملPopulation rate codes carried by mean, fluctuation and synchrony of neuronal firings
A population of firing neurons is expected to carry information not only by mean firing rate but also by fluctuation and synchrony among neurons. In order to examine this possibility, we have studied responses of neuronal ensembles to three kinds of inputs: mean-, fluctuation-and synchrony-driven inputs. The generalized rate-code model including additive and multiplica-tive noise (H. Hasegawa, ...
متن کاملResponses of neuronal ensembles to mean-driven, fluctuation-driven and synchrony-driven rate-code inputs
Equations of motion for mean firing rate, fluctuation and synchrony have been derived with the use of the augmented moment method (AMM) for the generalized rate-code model including additive and multiplicative noise (H. Hasegawa, Phys. Rev. E 75, 051904 (2007)). We have investigated the response of a neuronal ensemble to three kinds of inputs: mean-driven, fluctuation-driven and synchrony-drive...
متن کاملAnalysis and Synthesis of Facial Expressions by Feature-Points Tracking and Deformable Model
Face expression recognition is useful for designing new interactive devices offering the possibility of new ways for human to interact with computer systems. In this paper we develop a facial expressions analysis and synthesis system. The analysis part of the system is based on the facial features extracted from facial feature points (FFP) in frontal image sequences. Selected facial feature poi...
متن کاملSelf-recognition in the perception of actions performed in synchrony with music.
This study investigated self-recognition in point-light displays depicting actions performed in synchrony with music. Participants were recorded executing three different actions (dancing, walking, and clapping) and were subsequently required to identify the agent (self versus other) from point-light displays with or without the accompanying music. Results indicate that while recognition accura...
متن کامل