Deep Learning for Earth Observation (DL4EO) released Oriented-Det v0.2.0, a one-stage detector with a 3× decoded-rIoU checkpoint reaching 81.58% eval-val mAP50 on DOTA.
Public source
Publisher name
Public post
Oriented-Det v0.2.0 is out. Rotated FCOS joins the zoo as the balanced, anchor-free one-stage detector. The 3× decoded-rIoU checkpoint reaches 81.58% eval-val mAP50 on D…
Company
Deep Learning for Earth Observation (DL4EO)
Turning satellite imagery into valuable information using data‑centric AI tools for real‑world Earth solutions
- Industry
- Information Technology & Services
- Company size
- 1 employee
About Deep Learning for Earth Observation (DL4EO)
DL4EO builds and qualifies custom deep‑learning object detectors for satellite imagery, managing global DL projects including imagery acquisition, MLOps workflows and human-in-the-loop data labeling to continuously improve models and audit metrics for reliability.
See moreLatest activity
Latest activity from Deep Learning for Earth Observation (DL4EO)
2 signals
Discover more
Similar signals
Similar public activity from other companies.
Products & Services
Roboflow
Roboflow released RFDETR 1.10.0, a training-focused Open Source release with a 10k-image dataset trained for 10 epochs on an L4 dataset going from 55 minutes to 31 minutes.
Products & Services
Lightly
Lightly released the new LT-DETRv2 model line-up for object detection and instance segmentation tasks with LightlyTrain, an open-source framework for self-supervised learning and downstream fine-tuning.
Products & Services
OpenCV
OpenCV implemented DEIMv2 with two backbone options, HGNetV2 and DINOv3, for the head detector.
Products & Services
Earth-i
Earth-i introduced FANTOM, a platform combining Earth Observation data, advanced analytics, machine learning, quality assurance, and map-based outputs to support agricultural and environmental land monitoring across entire countries.
Products & Services
Ultralytics