AI Illuminated
autor: The AI Illuminators
Nejnovější epizoda
MegaSaM: Accurate, Fast, and Robust Structure and Motion from Casual…
MegaSaM: Accurate, Fast, and Robust Structure and Motion from Casual…
O tomto podcastu
A new way to keep up with AI research. Delivered to your ears. Illuminated by AI.
Part of the GenAI4Good initiative.
Epizody
Nejnovějších 20 z 25 epizod
- MegaSaM: Accurate, Fast, and Robust Structure and Motion from Casual Dynamic Videos · 7 min · Řada 1 Epizoda 250:00 Introduction 0:20 Limitations of traditional SfM and SLAM techniques. 0:57 Shortcomings of existing neural network methods. 1:07 MegaSaM's approach: balance of accuracy…
- SVDQuant: Absorbing Outliers by Low-Rank Components for 4-Bit Diffusion Models · 7 min · Řada 1 Epizoda 24[00:00] SVD-Quant: 4-bit diffusion model quantization [00:27] Challenge: Outlier sensitivity in 4-bit quantization [00:59] Solution: Smoothing + SVD approach [01:37]…
- D3RoMa: Disparity Diffusion-based Depth Sensing for Material-Agnostic Robotic Manipulation · 15 min · Řada 1 Epizoda 23[00:00] Intro [00:18] Current limitations in depth-sensing technology [00:56] D3RoMa's diffusion model approach to depth estimation [01:47] Integration of geometric…
- Scaling Proprioceptive-Visual Learning with Heterogeneous Pre-trained Transformers · 11 min · Řada 1 Epizoda 22[00:00] Intro [00:21] Key problem: Poor generalization in robotic learning [00:51] HPT: New transformer architecture for robotics [00:59] Core components of HPT architecture…
- HOVER: Versatile Neural Whole-Body Controller for Humanoid Robots · 7 min · Řada 1 Epizoda 21[00:00] Introduction to Hover: Neural Whole Body Controller for Humanoids [00:15] Problem: Current controllers lack versatility across tasks [00:50] Human motion imitation as a…
- Precise and Dexterous Robotic Manipulation via Human-in-the-Loop Reinforcement Learning · 9 min · Řada 1 Epizoda 20[00:00] Intro [00:24] Tackles RL challenges using a visual backbone, efficient RL, and human feedback. [01:20] Pretrained backbone boosts stability and exploration efficiency…
- Local Policies Enable Zero-shot Long Horizon Manipulation · 8 min · Řada 1 Epizoda 19[00:00] Paper intro: Zero-shot robotic manipulation via local policies [00:26] Key challenges: Limited generalization and sim-to-real transfer [01:03] Local policies: Task…
- MENTOR: Mixture-of-Experts Network with Task-Oriented Perturbation for Visual Reinforcement Learning · 9 min · Řada 1 Epizoda 18[00:00] Introduction to Mentor system for visual RL [00:29] Problem: Sample inefficiency in robotic learning [00:59] Innovation: Mixture of Experts (MoE) architecture [01:55]…
- SkillMimicGen: Automated Demonstration Generation for Efficient Skill Learning and Deployment · 9 min · Řada 1 Epizoda 17[00:00] SkillGen: AI system for robotic learning and automation. [00:19] Core Innovation: Automated dataset generation from minimal human input. [01:12] Skill Segmentation…
- LVSM: A Large View Synthesis Model with Minimal 3D Inductive Bias · 7 min · Řada 1 Epizoda 16[00:00] Intro to LVSM: Novel transformer for view synthesis [00:14] Problems with existing 3D synthesis methods [00:59] LVSM architecture: encoder-decoder vs decoder-only…
- Dynamic 3D Gaussian Tracking for Graph-Based Neural Dynamics Modeling · 16 min · Řada 1 Epizoda 15[00:00] Introduction to 3D Gaussian tracking for robotic manipulation [00:26] Limitations of current video prediction methods [01:11] Advantages of 3D Gaussian representation…
- SPIRE: Synergistic Planning, Imitation, and Reinforcement Learning for Long-Horizon Manipulation · 8 min · Řada 1 Epizoda 14[00:00] Introduction [00:20] Core limitations in robot manipulation: challenges with RL and IL [01:08] SPIRE's hybrid approach: combining task planning with learning methods…
- VILA-U: a Unified Foundation Model Integrating Visual Understanding and Generation · 10 min · Řada 1 Epizoda 13[00:00] VILA-U: A unified visual AI model [00:29] Problem: Inefficiency of separate visual modules [01:11] Vision tower: Novel quantization approach [02:09] Training strategy…
- CooHOI: Learning Cooperative Human-Object Interaction with Manipulated Object Dynamics · 5 min · Řada 1 Epizoda 12[00:00] Intro [00:31] Challenge: Limited multi-humanoid training data [00:55] CooHOI's two-phase learning framework [01:49] Object dynamics as implicit agent communication…
- SynFlowNet: Design of Diverse and Novel Molecules with Synthesis Constraints · 7 min · Řada 1 Epizoda 11[00:00] Introduction to SynFlowNet [00:29] Problem: AI-generated molecules often can't be synthesized [01:17] Solution: SynFlowNet - uses real chemical reactions [02:03]…
- L3DG: Latent 3D Gaussian Diffusion · 18 min · Řada 1 Epizoda 10[00:00] Intro to L3DG for 3D modeling [00:32] Solving room-sized 3D scene complexity [01:36] VQ-VAE compresses 3D Gaussian representation [02:41] Generative sparse transpose…
- The Ingredients for Robotic Diffusion Transformers · 8 min · Řada 1 Epizoda 9[00:00] Intro [00:33] Combining transformers & diffusion models [01:12] Key design: Scalable attention blocks (AdaLN) [02:30] Efficient observation tokenization [03:45] DiT…
- Estimating Body and Hand Motion in an Ego-sensed World · 16 min · Řada 1 Epizoda 8[00:00] Introduction to EgoAllo system [00:38] Challenges in egocentric motion estimation [01:20] Importance of spatial/temporal invariance [02:11] Comparison of conditioning…
- Janus: Decoupling Visual Encoding for Unified Multimodal Understanding and Generation · 9 min · Řada 1 Epizoda 7[00:00] Intro [00:28] Limitation of existing unified models [00:57] Janus's decoupled visual encoding solution [01:18] Advantages of decoupling [02:03] Janus architecture…
- One Step Diffusion via Shortcut Models · 5 min · Řada 1 Epizoda 6[00:00] Introduction [00:23] Computational cost of traditional diffusion models [00:59] Reducing iterations in image generation [01:06] Shortcut models [01:39] Training…
