A robotic drummer

  • Thread starter Thread starter pinball1970
  • Start date Start date
AI Thread Summary
The discussion critiques a paper on a robotic drummer that claims to achieve high-precision drumming across various songs. Participants express skepticism about the robot's performance, noting issues such as the absence of a bass drum, incorrect rhythms, and timing problems. The animation depicting the drummer is seen as misleading, with concerns that it does not accurately represent the music. Critics argue that the claims of "high-precision drumming" are overstated, suggesting that the robot's learning process, based on MIDI transcriptions, may not reflect true musical understanding. The conversation highlights a broader concern about the tendency to oversell AI and robotics capabilities, calling for more transparency and honesty in research claims.
pinball1970
Gold Member
Messages
3,500
Reaction score
5,480
TL;DR Summary
In this paper, we introduce Robot Drummer, a humanoid capable of expressive, high-precision drumming across a diverse repertoire of songs. We formulate humanoid drumming as sequential fulfillment of timed contacts and transform drum scores into a Rhythmic Contact Chain. To handle the long-horizon nature of musical performance, we decompose each piece into fixed-length segments and train a single policy across all segments in parallel using reinforcement learning.
Paper https://arxiv.org/abs/2507.11498

Article https://techxplore.com/news/2025-08-robotic-drummer-gradually-human-behaviors.html

Not my area of expertise at all, Robotics, programming but I do play drums. If you play the clip there is a lot wrong the with basic things the drummer is playing.
In the animation the drummer is stood, there no bass drum, is playing open hand - not usual.
If the animation is purely for illustration and not representative of the music then we can look at that also.
I cannot hear a separate BD, sounds like backing track.
Goes out of time.
Wrong rhythm and also goes "upside" down in Roxanne.

So I disagree with the abstract "high-precision drumming across a diverse repertoire of songs."

Opinion on what they have actually done from our high tech members will be welcome and also musicians who may be a little confused with the claims and results.
 
  • Like
  • Informative
Likes phinds, berkeman and Greg Bernhardt
Computer science news on Phys.org
I think they have some catching up to do... :wink:



1754670917310.webp
 
  • Wow
  • Like
Likes Greg Bernhardt and pinball1970
pinball1970 said:
Opinion on what they have actually done from our high tech members will be welcome and also musicians who may be a little confused with the claims and results.

My opinion is not printable within forum rules.

Whoever has published this should have their funding withdrawn.
 
Last edited:
  • Agree
  • Like
Likes 256bits and pinball1970
pinball1970 said:
TL;DR Summary: In this paper, we introduce Robot Drummer, a humanoid capable of expressive, high-precision drumming across a diverse repertoire of songs. We formulate humanoid drumming as sequential fulfillment of timed contacts and transform drum scores into a Rhythmic Contact Chain. To handle the long-horizon nature of musical performance, we decompose each piece into fixed-length segments and train a single policy across all segments in parallel using reinforcement learning.

So I disagree with the abstract "high-precision drumming across a diverse repertoire of songs."
From the article:
They found that the robot effectively learned complex rhythmic structures and played the songs with high accuracy, often with a rhythmic precision of over 90%.

So if the robo improves, then it becomes what "extreme ultra high precision" gobblylygook.
The authors of this 'study', and pretty much anything associated with AI, or robotics, fell into the present day common theme of overselling inferiority and defective products or research.

"To address this gap, we introduce Robot Drummer, a general RL framework for humanoid drumming that learns to perform drum patterns derived from MIDI transcriptions."
That's what I thought - Midi drumming, as the simulation video sounds so awful x86 16 bit, which is grating to the ears.

And only for the reason of over hype should this be cancelled. If they can't be truthful, then when will they ever. I imagine a lot went into this project to have the simulated robo read 'sheet' music, which is quite an accomplishment. Emphasis on THAT would be more in accordance with Machine Learning and the algorithms that went into it. But the article is sparse of details, and one cannot say if the robo is actually reading music on its own, or just flopping its arms around from directions given by a non AI program.
 
In my discussions elsewhere, I've noticed a lot of disagreement regarding AI. A question that comes up is, "Is AI hype?" Unfortunately, when this question is asked, the one asking, as far as I can tell, may mean one of three things which can lead to lots of confusion. I'll list them out now for clarity. 1. Can AI do everything a human can do and how close are we to that? 2. Are corporations and governments using the promise of AI to gain more power for themselves? 3. Are AI and transhumans...
Thread 'ChatGPT Examples, Good and Bad'
I've been experimenting with ChatGPT. Some results are good, some very very bad. I think examples can help expose the properties of this AI. Maybe you can post some of your favorite examples and tell us what they reveal about the properties of this AI. (I had problems with copy/paste of text and formatting, so I'm posting my examples as screen shots. That is a promising start. :smile: But then I provided values V=1, R1=1, R2=2, R3=3 and asked for the value of I. At first, it said...

Similar threads

Replies
13
Views
4K
Replies
1
Views
3K
Replies
1
Views
10K
Back
Top