AI-Based On-Board Maritime Object Detection for Earth Observation Payload Data Reduction on Versal Embedded Hardware
We study onboard vessel detection as a way to select what is downlinked, which reduces the data according to its content rather than coding every pixel; it is complementary to conventional onboard compression.
ProofPaper ↗
Key points
- Very-high-resolution Earth-observation satellites acquire more data than they can store and downlink, while in maritime surveillance the vessels cover a tiny fraction of each scene.
- On our dense harbor and coastal scenes, tiles keep 98% of the vessels with 29% of the scene volume, and crops 83% with 3%.
Sources (1)
- [1]AI-Based On-Board Maritime Object Detection for Earth Observation Payload Data Reduction on Versal Embedded HardwarearXiv (AI, ML, NLP, CV, robotics, multi-agent) · Oct 8, 03:48 PM
We study onboard vessel detection as a way to select what is downlinked, which reduces the data according to its content rather than coding every pixel; it is complementary to conventional onboard compression.
Very-high-resolution Earth-observation satellites acquire more data than they can store and downlink, while in maritime surveillance the vessels cover a tiny fraction of each scene.
Extractive summary: sentences quoted from the sources.
Before this
- Oct 8, 2026Deflating the Hessian: Rank-4 W4A4 Quantization for Multimodal Diffusion Transformers
- Oct 8, 2026Bridging KV-Cache Quantization and Linear Attention: From Theory to Pretrained Weight Migration
- Oct 8, 2026ConwayResearch/Underdog-Saluki-27B-1.0
- Oct 7, 2026Cache the Encoder Within:Compact, Reusable Memory across LLM Queries
- Oct 7, 2026TR-PTQ: High-Accuracy Integer-Only Transformer Post Training Quantization via Taylor Region Reformulation
- Oct 6, 2026mw00/project-maya: Project Maya - run GLM-5.3-Flash (321B MoE) privately on your own GPU(s), from one card to 16, NVIDIA or AMD, Linux or Windows. Engine, serv