Today, Manca, Luka and I presented the Compositional hierarchical model at the Music of Science session at Music Tech Fest in Ljubljana. As a part of his undergraduate degree, Luka built an awesome transcription editor which uses the CHM as classifier. It will probably never reach version 1.0 since the “ideas” stack grows exponentially. However, we are […]
We finally started the implementation of the singing stairs. The proximity sensors (for tracking the space along the bridge) are in place! Oh, and we switched to the RasPI2 in the meanwhile
During the last two days we presented two prominent research studies – the compositional hierarchical model for music information retrieval (including its applications to the automated chord estimation, the multiple fundamental frequency estimation and the mood estimation tasks), and the Moodo dataset – a large-scale dataset connecting user responses to color, music and emotions. Both papers […]
Today, the music part of LGM lab presented 3 papers at ERK 2014 conference at Portorož. Manca presented her work with the compositional hierarchical model for multiple fundamental frequency estimation; Primož presented his work on gathering a large-scale dataset including the mood responses to music and Matevž presented his work on evaluating the robustness of the compositional hierarchical […]
On Saturday we presented our novel research in using the compositional hierarchical model for MIR on symbolic data for pattern recognition. Although the majority of sessions were music theory related, we presented at Computational music analysis session, hosted by Dave Meredith. The abstract of the presentation (can be found here) is posted below. More pictures […]
Today, we (Grega) are presenting at SweCog 2014 Conference in Lund. We are presenting the analysis of the dataset from the cognitive perspective, as well as the relations between the modalities (music and color). An original file is available here: Poster_file. We are shortly planning on presenting an extensive MIR analysis at ISMIR 2014 in Taipei.
We are currently at IWSSIP conference which is this year held in Dubrovnik! We have presented our stacking approach, extending the previous research to the usage of SVMs and refined compositional features! Feel free to ping us for a cup of coffee!
Kakšno glasbo poslušaš ko si dobre volje? Kaj pa takrat, ko si napet in ti gredo vsi na živce? Ali vidiš rdeče pred očmi, ko si jezen? Se kdaj počutiš »modro«? Tudi ti pozeleniš od ljubosumja? Kaj pa tvoji prijatelji in znanci, dojemajo opisane stvari na isti način? Na ta vprašanja želimo odgovoriti v raziskavi […]
We’ve successfully submitted a paper considering the visualization of the compositional hierarchical model for deeper music analysis and music generation. The paper can be found here! (Visualization of a Deep Architecture using the Compositional Hierarchical Model). This is a preview of the proposed visualisation We are working on an extended version of the visualization […]