Multi-Scale
متعدد المقاييس
معالجة المعلومات على مستويات تقريب مختلفة في آنٍ واحد لالتقاط الأجسام الصغيرة والكبيرة معاً.
Processing information at multiple zoom levels simultaneously to capture both small and large objects.
Also translated asعبر المقاييس
First appears in this corpus in: Pixel Recurrent Neural Networks (2016)
Appears in these papers
- DeepLab: Semantic Image Segmentation with Deep Convolutional Nets, Atrous Convolution, and Fully Connected CRFs2017in the sky ✦
- Feature Pyramid Networks for Object Detection2017in the sky ✦
- Grounding DINO: Marrying DINO with Grounded Pre-Training for Open-Set Object Detection2023in the sky ✦
- Grounding DINO: Marrying DINO with Grounded Pre-Training for Open-Set Object Detection2023in the sky ✦
- Pixel Recurrent Neural Networks2016in the sky ✦
- SegFormer: Simple and Efficient Design for Semantic Segmentation with Transformers2021in the sky ✦
- SSD: Single Shot MultiBox Detector2016in the sky ✦
- YOLOv3: An Incremental Improvement2018in the sky ✦
- YOLOv3: An Incremental Improvement2018in the sky ✦