Twitter | Pretraživanje | |
Jonathan A. Michaels
time! How do we generate the right muscle commands to grasp objects? We present a neural network model that replicates the vision to action pipeline for grasping objects and shows internal activity very similar to the monkey brain.
Reply Retweet Označi sa "sviđa mi se" More
Jonathan A. Michaels 25. kol
Odgovor korisniku/ci @JonAMichaels
Monkeys grasped and lifted many objects while we recorded neural activity in the grasping circuit (AIP, F5, & M1 - see original paper ). All of these areas have been shown to be necessary for properly pre-shaping the hand during grasping.
Reply Retweet Označi sa "sviđa mi se"
Jonathan A. Michaels 25. kol
Odgovor korisniku/ci @JonAMichaels
We show that the advanced layers of a convolutional neural network trained to identify objects (Alexnet) has features very similar to those in AIP, and may therefore be reasonable inputs to the grasping circuit, while muscle velocity was most congruent with activity in M1.
Reply Retweet Označi sa "sviđa mi se"
Jonathan A. Michaels 25. kol
Odgovor korisniku/ci @JonAMichaels
Based on these results, we constructed a modular neural network model (top of thread) to transform visual images of objects into the muscle kinematics necessary to grasp them. Activity in the model was very similar to neural activity while monkeys completed the same task.
Reply Retweet Označi sa "sviđa mi se"
Jonathan A. Michaels 25. kol
Odgovor korisniku/ci @JonAMichaels
We tested how different neural network architectures and regularizations affected these results, finding that modular networks with visual input best matched neural data and showed similar inter-area relationships as in the brain.
Reply Retweet Označi sa "sviđa mi se"
Jonathan A. Michaels 25. kol
Odgovor korisniku/ci @JonAMichaels
Importantly, networks used simple computational strategies for maintaining, reorganizing, and executing movements, relying on a single fixed point during memory and a single fixed point during movement.
Reply Retweet Označi sa "sviđa mi se"
Jonathan A. Michaels 25. kol
Odgovor korisniku/ci @JonAMichaels
This simple strategy allowed networks to generalize well to novel objects, even predicting the real neural activity for these objects, providing a powerful predictive model of how flexible grasp control may be implemented in the primate brain!
Reply Retweet Označi sa "sviđa mi se"
Jonathan A. Michaels 25. kol
Odgovor korisniku/ci @SSchaffelhofer
Thanks to my fantastic collaborators , Andres Agudelo-Toro, Hans Scherberger.
Reply Retweet Označi sa "sviđa mi se"
Jonathan A. Michaels 25. kol
Odgovor korisniku/ci @DPZ_eu @shenoystanford @andpru
Thanks to the for supporting the work, and and for letting me keep working on the paper when I had more important things to do.
Reply Retweet Označi sa "sviđa mi se"
Jonathan A. Michaels 25. kol
Odgovor korisniku/ci @dann_benjamin @bsauerbrei1 i 4 ostali
Thanks to everyone who gave great feedback along the way, including , , , , , , and many others.
Reply Retweet Označi sa "sviđa mi se"
Mohsen Omrani 25. kol
Odgovor korisniku/ci @JonAMichaels
Jonathan, while preaching the nested feedback loops, you forgot to cite an important paper providing evidence for it happening in the motor system ;)
Reply Retweet Označi sa "sviđa mi se"
Jonathan A. Michaels 25. kol
Odgovor korisniku/ci @Mohsen_Omrani
Sorry!! I cite your paper many times in all my grant applications :).
Reply Retweet Označi sa "sviđa mi se"