Squeezing nnU-Nets with Knowledge Distillation for On-Board Cloud Detection

• URL: http://arxiv.org/abs/2306.09886v1
• Date: Fri, 16 Jun 2023 14:53:36 GMT
• Title: Squeezing nnU-Nets with Knowledge Distillation for On-Board Cloud Detection
• Authors: Bartosz Grabowski, Maciej Ziaja, Michal Kawulok, Piotr Bosowski, Nicolas Long\'ep\'e, Bertrand Le Saux, Jakub Nalepa
• Abstract summary: nnU-Nets is a self-reconfigurable framework able to perform meta-learning of a segmentation network over various datasets. We compress nnU-Nets with knowledge distillation into much smaller and compact U-Nets. Our approach was ranked within the top 7% best solutions (across 847 teams) in the On Cloud N: Cloud Cover Detection Challenge.
• Score: 29.014110832117993
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Cloud detection is a pivotal satellite image pre-processing step that can be performed both on the ground and on board a satellite to tag useful images. In the latter case, it can reduce the amount of data to downlink by pruning the cloudy areas, or to make a satellite more autonomous through data-driven acquisition re-scheduling. We approach this task with nnU-Nets, a self-reconfigurable framework able to perform meta-learning of a segmentation network over various datasets. Unfortunately, such models are commonly memory-inefficient due to their (very) large architectures. To benefit from them in on-board processing, we compress nnU-Nets with knowledge distillation into much smaller and compact U-Nets. Our experiments, performed over Sentinel-2 and Landsat-8 images revealed that nnU-Nets deliver state-of-the-art performance without any manual design. Our approach was ranked within the top 7% best solutions (across 847 teams) in the On Cloud N: Cloud Cover Detection Challenge, where we reached the Jaccard index of 0.882 over more than 10k unseen Sentinel-2 images (the winners obtained 0.897, the baseline U-Net with the ResNet-34 backbone: 0.817, and the classic Sentinel-2 image thresholding: 0.652). Finally, we showed that knowledge distillation enables to elaborate dramatically smaller (almost 280x) U-Nets when compared to nnU-Nets while still maintaining their segmentation capabilities.

Related papers

• Remote Sensing Image Classification Using Convolutional Neural Network (CNN) and Transfer Learning Techniques [1.024113475677323]
  This study investigates the classification of aerial images depicting transmission towers, forests, farmland, and mountains. To complete the classification job, features are extracted from input photos using a Convolutional Neural Network (CNN) architecture. Our study shows that transfer learning models and MobileNetV2 in particular, work well for landscape categorization.
  arXiv  Detail & Related papers  (2025-03-04T11:19:18Z)
• Semantic Segmentation in Satellite Hyperspectral Imagery by Deep Learning [54.094272065609815]
  We propose a lightweight 1D-CNN model, 1D-Justo-LiuNet, which outperforms state-of-the-art models in the hypespectral domain. 1D-Justo-LiuNet achieves the highest accuracy (0.93) with the smallest model size (4,563 parameters) among all tested models.
  arXiv  Detail & Related papers  (2023-10-24T21:57:59Z)
• An Accurate and Efficient Neural Network for OCTA Vessel Segmentation and a New Dataset [2.8743451550676866]
  We propose an accurate and efficient neural network for retinal vessel segmentation in OCTA images. The proposed network achieves accuracy comparable to other SOTA methods, while having fewer parameters and faster inference speed. We create a new dataset containing 918 OCTA images and their corresponding vessel annotations.
  arXiv  Detail & Related papers  (2023-09-18T04:47:12Z)
• A Light-weight Deep Learning Model for Remote Sensing Image Classification [70.66164876551674]
  We present a high-performance and light-weight deep learning model for Remote Sensing Image Classification (RSIC) By conducting extensive experiments on the NWPU-RESISC45 benchmark, our proposed teacher-student models outperforms the state-of-the-art systems.
  arXiv  Detail & Related papers  (2023-02-25T09:02:01Z)
• ConvNeXt V2: Co-designing and Scaling ConvNets with Masked Autoencoders [104.05133094625137]
  We propose a fully convolutional masked autoencoder framework and a new Global Response Normalization layer. This co-design of self-supervised learning techniques and architectural improvement results in a new model family called ConvNeXt V2, which significantly improves the performance of pure ConvNets.
  arXiv  Detail & Related papers  (2023-01-02T18:59:31Z)
• MogaNet: Multi-order Gated Aggregation Network [64.16774341908365]
  We propose a new family of modern ConvNets, dubbed MogaNet, for discriminative visual representation learning. MogaNet encapsulates conceptually simple yet effective convolutions and gated aggregation into a compact module. MogaNet exhibits great scalability, impressive efficiency of parameters, and competitive performance compared to state-of-the-art ViTs and ConvNets on ImageNet.
  arXiv  Detail & Related papers  (2022-11-07T04:31:17Z)
• Self-Configuring nnU-Nets Detect Clouds in Satellite Images [30.46904432868366]
  nnU-Nets is a self-reconfigurable framework able to perform meta-learning of a segmentation network over various datasets. Our experiments, performed over Sentinel-2 and Landsat-8 multispectral images revealed that nnU-Nets deliver state-of-the-art cloud segmentation performance without any manual design.
  arXiv  Detail & Related papers  (2022-10-24T23:39:58Z)
• EdgeNeXt: Efficiently Amalgamated CNN-Transformer Architecture for Mobile Vision Applications [68.35683849098105]
  We introduce split depth-wise transpose attention (SDTA) encoder that splits input tensors into multiple channel groups. Our EdgeNeXt model with 1.3M parameters achieves 71.2% top-1 accuracy on ImageNet-1K. Our EdgeNeXt model with 5.6M parameters achieves 79.4% top-1 accuracy on ImageNet-1K.
  arXiv  Detail & Related papers  (2022-06-21T17:59:56Z)
• Rethinking BiSeNet For Real-time Semantic Segmentation [6.622485130017622]
  BiSeNet has been proved to be a popular two-stream network for real-time segmentation. We propose a novel structure named Short-Term Dense Concatenate network (STDC) by removing structure redundancy.
  arXiv  Detail & Related papers  (2021-04-27T13:49:47Z)
• U-Net Based Architecture for an Improved Multiresolution Segmentation in Medical Images [0.0]
  We have proposed a fully convolutional neural network for image segmentation in a multi-resolution framework. In the proposed architecture (mrU-Net), the input image and its down-sampled versions were used as the network inputs. We trained and tested the network on four different medical datasets.
  arXiv  Detail & Related papers  (2020-07-16T10:19:01Z)
• Road Segmentation on low resolution Lidar point clouds for autonomous vehicles [3.6020689500145653]
  We evaluate the effect of subsampling image based representations of dense point clouds on the accuracy of the road segmentation task. We introduce the usage of the local normal vector with the LIDAR's spherical coordinates as an input channel to existing LoDNN architectures.
  arXiv  Detail & Related papers  (2020-05-27T00:38:39Z)
• Improved Residual Networks for Image and Video Recognition [98.10703825716142]
  Residual networks (ResNets) represent a powerful type of convolutional neural network (CNN) architecture. We show consistent improvements in accuracy and learning convergence over the baseline. Our proposed approach allows us to train extremely deep networks, while the baseline shows severe optimization issues.
  arXiv  Detail & Related papers  (2020-04-10T11:09:50Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.