Proud of Mehdi's successful PhD presentation 🎓 of his work combining #ReinforcementLearning and #FormalMethods for robots to learn online Automatic Symbolic Goal Abstraction via Reachability Analysis in Hierarchical Reinforcement Learning
📽️ youtu.be/vOjRKN8MFAo
#EmergentSymbols #RobotLearning

I finally uploaded my presentation "The #IntrinsicMotivation of #ReinforcementLearning and #ImitationLearning for sequential tasks" which overviews my past 10 years of research
📽️ youtu.be/tPndN5dntoQ
#ActiveImitationLearning #CompositionalLearning #ActivityRecognition #ADL #RobotLearning #RobotCoach

Proud that Qi will present in 2D human pose estimation #HPE at #CVsports @cvprconference.bsky.social how his pose prior-guided refinement approach based on Neural Distance Fields,
polar coordinate representation and its metric, tackles data scarcity of long-jump videos
👉 arxiv.org/abs/2505.03445

A symbolic representation of tasks is key for compositionality. Our goal-conditioned hierarchical #ReinforcementLearning algo STAR learns online a discrete representation of continuous sensorimotor space #ReachabilityAnalysis #ContinualLearning #ICLR2024 @iclr-conf.bsky.social
📽️ youtu.be/H5-ufMj2AE0

For intelligent tutoring systems, Louis presented @ieee-icdl.bsky.social 2023 how to learn the knowledge structure by uncovering prerequisite relations from students' logs, how #IntrinsicMotivation derives a personalized curriculum
👉 inria.hal.science/hal-04401880v1
📽️ www.youtube.com/watch?v=QNRq...

Our medical dataset of low-back pain patients is published at #IJCNN2024
📣 The 1st #MovementAnalysis dataset for #PhysicalRehabilitation with daily recordings of patients follow-up throughout a month-long clinical care #RobotPoppy

⬇️ nguyensmai.free.fr/KeraalDatase...
👩‍💻 github.com/nguyensmai/K...

Louis presented our test of #DeepLearning unpaired translation to solve the correspondence problem for imitation of patients with different biomechanics by any #humanoidRobot without labelled data @HRI 2024 Conference
👉 hal.science/hal-04401885v1

📽️ youtu.be/I27z7kqpyyc

At the #ML4ITS2023 Workshop at @ecmlpkdd.org, Damien describes how he uses language models and #GPT architecture to obtain a contextualized representation of Activities of Daily Living #ADL for classification and transfer learning
👉 inria.hal.science/hal-04858744v1
#LLM #selfSupervisedLearning

So excited !
📣medical report & patient-therapist evaluation of the #RobotCoach #RobotPoppy for #PhysicalRehabilitation in clinical trial 🏥
with @mdevanne.bsky.social is published online
👉 onlinelibrary.wiley.com/doi/10.1155/...

www.youtube.com/watch?v=db1X...

How can knowledge from easier tasks be transferred to complex tasks while the robot learns tasks relationships? How can it discover new learnable subgoals for more complex tasks? How can human demonstrations be beneficial to an active learning robot?

👉 hal.science/hal-03157219