11 papers
Finding Change in Satellite Archives from Text: How to Combine Before-and-After Images Efficiently
Simon Roy, Mark Bong, Giovanni Beltrame
Operational Earth observation increasingly calls for answering queries such as ``find the image pairs where a new building appeared.'' This means searching an archive of before-and…
VISTA: Scale-Aware Visual Navigation via Action History Conditioning
Maeva Guerrier, Koki Kobayashi, Simon Roy +2
Vision Navigation Foundation Models (VNMs) promise end-to-end learned navigation policies capable of zero-shot deployment across diverse embodiments and environments. To maintain g…
Latent Goal Prediction from Language for Model-Based Planning
Samuel Barbeau, Simon Roy, Giovanni Beltrame +2
Joint-Embedding Predictive Architectures (JEPAs) enable agents to plan in latent space by imagining the outcomes of candidate actions, yet task specification remains a bottleneck.…
Revisiting the Learning Objectives of Vision-Language Reward Models
Simon Roy, Samuel Barbeau, Giovanni Beltrame +2
Learning generalizable reward functions is a core challenge in embodied intelligence. Recent work leverages contrastive vision language models (VLMs) to obtain dense, domain-agnost…
BitMar: Low-Bit Multimodal Fusion with Episodic Memory for Edge Devices
Euhid Aman, Esteban Carlin, Hsing-Kuo Pao +3
Cross-attention transformers and other multimodal vision-language models excel at grounding and generation; however, their extensive, full-precision backbones make it challenging t…
Frequency-based View Selection in Gaussian Splatting Reconstruction
Monica M. Q. Li, Pierre-Yves Lajoie, Giovanni Beltrame
Three-dimensional reconstruction is a fundamental problem in robotics perception. We examine the problem of active view selection to perform 3D Gaussian Splatting reconstructions w…
Variable Time Step Reinforcement Learning for Robotic Applications
Dong Wang, Giovanni Beltrame
Traditional reinforcement learning (RL) generates discrete control policies, assigning one action per cycle. These policies are usually implemented as in a fixed-frequency control…
MOSEAC: Streamlined Variable Time Step Reinforcement Learning
Dong Wang, Giovanni Beltrame
Traditional reinforcement learning (RL) methods typically employ a fixed control loop, where each cycle corresponds to an action. This rigidity poses challenges in practical applic…
Reinforcement Learning with Elastic Time Steps
Dong Wang, Giovanni Beltrame
Traditional Reinforcement Learning (RL) policies are typically implemented with fixed control rates, often disregarding the impact of control rate selection. This can lead to ineff…
Deployable Reinforcement Learning with Variable Control Rate
Dong Wang, Giovanni Beltrame
Deploying controllers trained with Reinforcement Learning (RL) on real robots can be challenging: RL relies on agents' policies being modeled as Markov Decision Processes (MDPs), w…
Event-based RGB-D sensing with structured light
Seyed Ehsan Marjani Bajestani, Giovanni Beltrame
Event-based cameras (ECs) are bio-inspired sensors that asynchronously report brightness changes for each pixel. Due to their high dynamic range, pixel bandwidth, temporal resoluti…