analiza aiqbench
FLINT demonstrates that high-accuracy off-road navigation can be achieved with lightweight neural networks running on standard CPUs, eliminating the need for expensive multi-sensor stacks or heavy foundation models. By leveraging self-supervised learning from RGB data alone, this approach significantly reduces computational overhead while maintaining superior performance in real-world field trials compared to larger baselines.
FLINT: Fast Lightweight Inference for Traversability
arXiv:2609.26857v1 Announce Type: cross Abstract: Navigation in off-road conditions is challenging due to the lack of structure. There is no fixed vocabulary for what is traversable. The traversability depends on both the environment and the embodiment's dynamics. Neither of these two variables can be hand-labeled at scale. Thus, traversability has to be learned by the embodiment's own experience. Modern platforms tend to use multiple sensors to estimate traversability and navigate: RGBD cameras, lidar, radar, IMU, with computationally intensive platforms to run inference on neural networks. Against this trend, we propose FLINT, a lightweight traversability estimator: a 21.6M-parameter backbone, 38\times smaller than a comparable foundation-model backbone, that scores higher on held-out terrain probes and runs at 14.7 FPS on CPU alone using a RGB camera has the only sensor. Despite that gap in scale, FLINT produces a cheaper, more accurate costmap than a deployed foundation-model system (WildOS) on 23 of 24 replayed field logs. We compare different self-supervised learning signals and deploy the resulting models on a real platform in closed-loop field trials: the best self-supervised head reaches 99% autonomy over the route, outperforming a human-label-trained baseline deployed live on the same course. Our results show that heavy sensing and computing are not necessary for traversability estimation.
Abstract:Navigation in off-road conditions is challenging due to the lack of structure. There is no fixed vocabulary for what is traversable. The traversability depends on both the environment and the embodiment's dynamics. Neither of these two variables can be hand-labeled at scale. Thus, traversability has to be learned by the embodiment's own experience. Modern platforms tend to use multiple sensors to estimate traversability and navigate: RGBD cameras, lidar, radar, IMU, with computationally intensive platforms to run inference on neural networks. Against this trend, we propose FLINT, a lightweight traversability estimator: a 21.6M-parameter backbone, 38\times smaller than a comparable foundation-model backbone, that scores higher on held-out terrain probes and runs at 14.7 FPS on CPU alone using a RGB camera has the only sensor. Despite that gap in scale, FLINT produces a cheaper, more accurate costmap than a deployed foundation-model system (WildOS) on 23 of 24 replayed field logs. We compare different self-supervised learning signals and deploy the resulting models on a real platform in closed-loop field trials: the best self-supervised head reaches 99% autonomy over the route, outperforming a human-label-trained baseline deployed live on the same course. Our results show that heavy sensing and computing are not necessary for traversability estimation.
Robotics (cs.RO); Artificial Intelligence (cs.AI)
arXiv:2609.26857 [cs.RO]
(or arXiv:2609.26857v1 [cs.RO] for this version)
https://doi.org/10.48550/arXiv.2609.26857 Focus to learn more arXiv-issued DOI via DataCite
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Text integral sub licența deschisă a publicației · 295 cuvinte
Subiecte conexe
Triton-Based Optimization of Video Sparse Attention on ROCm
AMD · 13 iul. 2026
Requests of a Feather Must Flock Together: Batch Size vs. Prefix Homogeneity in LLM Inference
arXiv · 14 oct. 2026
PIG: Parallel Inconsistency Guidance for Accelerating Diffusion Language Models
arXiv · 14 oct. 2026
mKernel: Fast Multi-GPU, Multi-Node Fused Kernels
arXiv · 16 sept. 2026