![ccrmalite1](/img/default-banner.jpg)
- 251
- 998 288
ccrmalite1
United States
Приєднався 5 чер 2010
Adaptive and interactive machine listening with minimal supervision
DSP Seminar - February 10, 2023. CCRMA, Stanford
Abstract: Nowadays deep learning-based approaches have become popular tools and achieved promising results in machine listening. However, a deep model that generalizes well needs to be trained on a large amount of labeled data. Rare, fine-grained, or newly emerged classes (e.g. a rare musical instrument or a new sound effect) where large-scale data collection is hard or simply impossible are often considered out-of-vocabulary and unsupported by machine listening systems. In this thesis work, we aim to provide new perspectives and approaches to machine listening tasks with limited labeled data. Specifically, we focus on algorithms that are designed to work with few labeled data (e.g. few-shot learning) and incorporate human input to guide the machine. The goal is to develop flexible and customizable machine listening systems that can adapt to different tasks in a data-efficient way with the help of minimal human intervention.
Bio: Yu Wang is a Ph.D. candidate in Music Technology at the Music and Audio Research Laboratory (MARL) at New York University, advised by Prof. Juan Pablo Bello. Her research interests focus on machine learning and signal processing for music and general audio. She has interned with Adobe Research, Spotify, and Google Magenta. Before joining MARL in 2017, she was in the Music Recording and Production program at the Institute of Audio Research. She holds two M.S. degrees in Materials Science & Engineering from Massachusetts Institute of Technology (2015) and National Taiwan University (NTU) (2012), and a B.S. in Physics from NTU (2010). Yu is a guitar player and also enjoys sound engineering. Japanese math rock is her current favorite music genre.
ccrma.stanford.edu/events/adaptive-and-interactive-machine-listening-with-minimal-supervision
Abstract: Nowadays deep learning-based approaches have become popular tools and achieved promising results in machine listening. However, a deep model that generalizes well needs to be trained on a large amount of labeled data. Rare, fine-grained, or newly emerged classes (e.g. a rare musical instrument or a new sound effect) where large-scale data collection is hard or simply impossible are often considered out-of-vocabulary and unsupported by machine listening systems. In this thesis work, we aim to provide new perspectives and approaches to machine listening tasks with limited labeled data. Specifically, we focus on algorithms that are designed to work with few labeled data (e.g. few-shot learning) and incorporate human input to guide the machine. The goal is to develop flexible and customizable machine listening systems that can adapt to different tasks in a data-efficient way with the help of minimal human intervention.
Bio: Yu Wang is a Ph.D. candidate in Music Technology at the Music and Audio Research Laboratory (MARL) at New York University, advised by Prof. Juan Pablo Bello. Her research interests focus on machine learning and signal processing for music and general audio. She has interned with Adobe Research, Spotify, and Google Magenta. Before joining MARL in 2017, she was in the Music Recording and Production program at the Institute of Audio Research. She holds two M.S. degrees in Materials Science & Engineering from Massachusetts Institute of Technology (2015) and National Taiwan University (NTU) (2012), and a B.S. in Physics from NTU (2010). Yu is a guitar player and also enjoys sound engineering. Japanese math rock is her current favorite music genre.
ccrma.stanford.edu/events/adaptive-and-interactive-machine-listening-with-minimal-supervision
Переглядів: 116
Відео
Meta-AF: Meta-Learning for Adaptive Filters
Переглядів 1,4 тис.Рік тому
DSP Seminar - November 18, 2022. CCRMA, Stanford Abstract: Adaptive filtering algorithms are pervasive throughout modern society and have had a significant impact on a wide variety of domains including audio processing, telecommunications, biomedical sensing, astrophysics and cosmology, seismology, and many more. Adaptive filters typically operate via specialized online, iterative optimization ...
Feedback Delay Networks for Artificial Reverberation - Sebastian Schlecht
Переглядів 2 тис.Рік тому
DSP Seminar - November 11, 2022. CCRMA, Stanford Abstract: Feedback delay networks (FDNs) are recursive filters widely used for artificial reverberation and decorrelation. While vast literature exists on a wide variety of reverb topologies, FDNs provide a unifying framework to design and analyze delay-based reverberators. This talk reviews recent advancements in the FDN theory, such as lossless...
DeepAFx-ST: Style Transfer of Audio Effects with Differentiable Signal Processing
Переглядів 1,2 тис.Рік тому
DSP Seminar - November 4, 2022. CCRMA, Stanford Abstract: We present a framework that can impose the audio effects and production style from one recording to another by example with the goal of simplifying the audio production process. We train a deep neural network to analyze an input recording and a style reference recording and predict the control parameters of audio effects used to render t...
CCRMA@HOME - September 4, 2020
Переглядів 1443 роки тому
CCRMAlites share new works from their homes into yours ... We hope you will consider supporting the causes below. Any amount helps! Kenneth Reams, an artist, writer, and activist currently held in solitary confinement in Arkansas. Alkimiya Transfer’s set features a new poem by Reams, entitled Distributive Justice. NAACP Legal Defense and Educational Fund Black Classical Music Educators Black Gi...
Quarantine Sessions #20
Переглядів 2383 роки тому
The Coronavirus Crisis has changed our lives, and we are in the midst of a long period without concerts as we knew them. In addition to the problem of large audiences, the regulations also make it 'virtually' impossible for musicians to get together, to rehearse or perform. However, many technologies and solutions are already available, helping us to find new ways of collaborating and transport...
Quarantine Sessions #19
Переглядів 823 роки тому
The Coronavirus Crisis has changed our lives, and we are in the midst of a long period without concerts as we knew them. In addition to the problem of large audiences, the regulations also make it 'virtually' impossible for musicians to get together, to rehearse or perform. However, many technologies and solutions are already available, helping us to find new ways of collaborating and transport...
Quarantine Sessions #18
Переглядів 1093 роки тому
The Coronavirus Crisis has changed our lives, and we are in the midst of a long period without concerts as we knew them. In addition to the problem of large audiences, the regulations also make it 'virtually' impossible for musicians to get together, to rehearse or perform. However, many technologies and solutions are already available, helping us to find new ways of collaborating and transport...
Quarantine Sessions #17
Переглядів 1193 роки тому
The Coronavirus Crisis has changed our lives, and we are in the midst of a long period without concerts as we knew them. In addition to the problem of large audiences, the regulations also make it 'virtually' impossible for musicians to get together, to rehearse or perform. However, many technologies and solutions are already available, helping us to find new ways of collaborating and transport...
Quarantine Sessions #16
Переглядів 963 роки тому
The Coronavirus Crisis has changed our lives, and we are in the midst of a long period without concerts as we knew them. In addition to the problem of large audiences, the regulations also make it 'virtually' impossible for musicians to get together, to rehearse or perform. However, many technologies and solutions are already available, helping us to find new ways of collaborating and transport...
Quarantine Sessions #15
Переглядів 2103 роки тому
The Coronavirus Crisis has changed our lives, and we are in the midst of a long period without concerts as we knew them. In addition to the problem of large audiences, the regulations also make it 'virtually' impossible for musicians to get together, to rehearse or perform. However, many technologies and solutions are already available, helping us to find new ways of collaborating and transport...
Quarantine Sessions #14
Переглядів 4544 роки тому
The Coronavirus Crisis has changed our lives, and we are in the midst of a long period without concerts as we knew them. In addition to the problem of large audiences, the regulations also make it 'virtually' impossible for musicians to get together, to rehearse or perform. However, many technologies and solutions are already available, helping us to find new ways of collaborating and transport...
Quarantine Sessions #13
Переглядів 1684 роки тому
The Coronavirus Crisis has changed our lives, and we are in the midst of a long period without concerts as we knew them. In addition to the problem of large audiences, the regulations also make it 'virtually' impossible for musicians to get together, to rehearse or perform. However, many technologies and solutions are already available, helping us to find new ways of collaborating and transport...
Audiovisual Meditation Journal - Julie Herndon
Переглядів 934 роки тому
During shelter-in-place, I recorded three audio scores meditations and sent them to pairs of artists, asking them to record either video or audio responses. The meditations focused on physical sensations and imagined states of being. My hope was to explore the strange sense of time and mental state that I and so many others experienced while in isolation: a reaching outward to community and a r...
Quarantine Sessions #5 - A Distributed Electroacoustic Network Improvisation
Переглядів 1794 роки тому
Quarantine Sessions #5 - A Distributed Electroacoustic Network Improvisation
Quarantine Sessions #4 - A Distributed Electroacoustic Network Improvisation
Переглядів 1134 роки тому
Quarantine Sessions #4 - A Distributed Electroacoustic Network Improvisation
Ada's Song: Teaching Machines to Sing - Patricia Alessandrini
Переглядів 3404 роки тому
Ada's Song: Teaching Machines to Sing - Patricia Alessandrini
Resurrections (2020) - Fernando Lopez-Lezcano (performance two)
Переглядів 1694 роки тому
Resurrections (2020) - Fernando Lopez-Lezcano (performance two)
Resurrections (2020) - Fernando Lopez-Lezcano (performance one)
Переглядів 7874 роки тому
Resurrections (2020) - Fernando Lopez-Lezcano (performance one)
IMSYS: SPECTRAL PANNING VIA FLOCKING ALGORITHMS IN A MULTI-CHANNEL SOUND ENVIRONMENT
Переглядів 1,1 тис.4 роки тому
IMSYS: SPECTRAL PANNING VIA FLOCKING ALGORITHMS IN A MULTI-CHANNEL SOUND ENVIRONMENT
Geir Davidsen and Matt Wright : Self-Playing Euphonium Mouthpiece
Переглядів 3065 років тому
Geir Davidsen and Matt Wright : Self-Playing Euphonium Mouthpiece
Να αξιωθώ νά ανάψω ενα κεράκι στην ΑΓΙΑ ΣΟΦΙΑ ΜΑΣ...νά ακούσω την Θεία Λειτουργία...κ ας αφήσω την τελευταία μου πνοή για τη χάρη του Θεού..
As a Minnesotan, "Born in Minneapolis but Grew up in St. Paul," is hilarious to me. To the locals this is ironically true. "You go to Minneapolis to sin and St. Paul to pray," as the old saying goes.
Immediately I don't dig the interviewer
Met Muhal around the time of this interview .feel privileged to have had that opportunity.
Why does 1 clap become 4 claps once it has arrived the A square ? I can understand if this ould be a diffuser delay or allpas , is it related ?
A is a 4x4 feedback matrix that specifies which inputs are propagated to which outputs. If A isn't an identity matrix, there's some intermixing going on. In the example shown at 9:35, the matrix seems to be non-zero for all elements, so all inputs input are propagated to all outputs.
One day
Lirycs of the song?
Hagia Sophia☦️once a church, always a church
Beautiful touching the soul
ΕΝΑΣ ΑΠΟ ΤΟΥΣ ΚΑΛΥΤΕΡΟΥΣ ΥΜΝΟΥΣ..ΧΕΡΟΥΒΙΚΟΣ ΥΜΝΟΣ..ΣΕ ΑΝΥΨΩΝΕΙ ΣΤΑ ΟΥΡΑΝΙΑ...ΝΑ ΝΑΙ ΕΥΛΟΓΗΜΕΝΟ !!!
Infinitely more beautiful than what they do there now. Lord forgive them.
hello this is Isabelle Reams , did not know you had done this video or posted that video about my husband, very happy you took the time to do this, :)
'Nevertheless', 'however', 'but' indicate that what is said or written next is a modifying thought ie modifies the initial statement by adding something unexpected or subtracting from the implications of the first statement. I am definitely missing something in all this stuff about categories. Seems to be complicating things unnecessarily. 'But' is an event in discourse space, he says. OK. I think this is a real mess of a lecture, it's all over the place. What's his point, what's his insight? Please, someone, enlighten me!
❤❤❤❤byzantin tradition❤❤❤❤
I also have friends at Stanford who fund it. Do they know your hacking Video Games Banana? Just wondered. How's Yuba City? Let's try to make a peace ok? thanks.
is amazing that this is free
Κάπως έτσι θα ακούγονται οι ψαλμωδίες των αγγέλων!
G E N I A L !!!
NO WAY SOMEONE PALYED IT
$10,000 is still real money
At 45:46, he talks about easy transposition and how it captures people's attention the most. Musical concepts aren't as difficult as the problem of understanding them in terms of every key equally clearly. The traditional keyboard is deceptively simple, and his xylophone like the Janko keyboard attempts to address the problem of its bias to C major, which every other scale has to be viewed as some distortion of. There's still no commercial 6-6 controller keyboard at anywhere near an equivalent price, for obvious reasons. Most people probably see the 6-row Janko keyboard as insanely complicated, yet if you created a keyboard based on the guitar fretboard, it would be far more complicated to a beginner than a Janko piano. (Actually, Starr Labs have created keyboards based on fretboards with strings tuned in 4ths). Nevertheless, I think those few working on alternative keyboards should simplify the design to something like a 3-row Reuther accordion (but as a full keyboard) rather than the 6-row Janko design. But the idea seems to appeal to those attracted to the design aspect rather than its practical application, and so the concept gets taken too far, and now the instrument has to embrace exotic tunings with each key independently tunable with lights, etc ...which would be fine if a simple, affordable 6-6 instrument was available for people to get to grips with first.
I bought a chromatone ct-312 for $500- I think they retail for a lot less than that (not even sure if/where they're still sold). But i'd say that's within the realm of affordability- certainly higher than your cheapest keyboard controllers though.
At 56 minutes in, the question about the augmented chord being applied in a similar way. Barry Harris didn't cover this, as far as I know, because it mainly concerns 3 note harmony. But, lowering or raising a note of each augmented triad gives the major & minor triads (respectively). So each augmented triad connects to 3 major & minor triads in a similar way to how each full diminished chord connects to 4 dominant 7ths. Lowering 2 augmented triad notes gives minor chords on the same root, whereas raising 2 augmented notes gives V+ to I progressions. I prefer the term 'full diminished' than 'diminished 7th', so we can say 'full augmented' to include the 8ve (as that divides it equally into three instead of four). So, lowering & raising notes of full augmented chords, (e.g. C E G# C), also gives the various chords, (e.g. C, Cmaj7, etc, or Am, A(#9), etc.
So beautiful. Thanks for sharing!
Proverbs are part of a typical mental status exam
Anybody know where I could find notation for this chant?
The fact that they played it with straight faces
This is actually really interesting. I mean, learning how to actually preform this must have been quite the feat
Dear god 😮
Is this a CONTRABASS CLARINET?
I feel sorry for the musicians. Fine for film . But as a concert, no way.
It's all in Roget's synopsis of categories (1852).
Wunderschön
My son has been studying/working/creating at CCRMA with this group and others using and expanding on this virtual acoustics technology. He and others are continuing making the Icons of Sound available acoustically to many many more to experience. CCRMA is at Stanford University.
this man has ascended to a higher level of musician than us mere mortals
Compositions like this one really only work in the live setting. This work is deeply profound with music to match the amazing performances and the perfect sets. It's just a wonderful opera. Still, well worth watching here on UA-cam.
Congratulations for keeping alive the ancient Eastern Byzantine Greek chants,,
Ancient=medical. Byzantine= estern Rome, ασχετε αμόρφωτε
@@panosant3960I think he is referring to the Greek language used in these hymns 😂
Amazing. Like going back in time.
Soooo cool 🙏🏽
Ahead of it’s time 👌🏽🔥
😮😮
I believe a dynamics range effect in this case should work both ways, so both as a compressor and an expander. Especially with the target demographic thats relying on a lot of vst instruments that lack dynamics
Beautiful sounds peaceful Angelic spiritually uplifting well done
Hauntingly beautiful.
I enjoyed this. It was interesting and not too much at once. Either it was unusually good or I have grown in my appreciation of new things in general.
Amazing!
"I have to go to school to pick up my backpack" kind of works because he has to go to THE school to get it, but since (we assume) it's a school day, if he steps into the school as is necessary, then he can't decently play truant in full view of everybody (the teachers, principal, what not) and walk out right after he's fetched the thing, so he is stuck and he has to go to school.
"FRENCHLANGUAGEXXX, quoi." means that you have successfully expressed in a succinct, practical way a complex thought. You have successfully boiled it down to its conceptual essence and practical import. It's a little nod to yourself for having achieved that, and an invitation for your interlocutor to acknowledge as much - while also admitting that some less important, but nonetheless existent, elements of the original thought have been lost through the boiling down process.
A Heritage.
Re the previous comment: UNfortunately Hagia Sofia is no longer a museum; it is now a mosque.