Lightweight, Pre-trained Transformers for Remote Sensing Timeseries
854efd16088ec0a2949395f17c9b60f9386aedb0ff5a819cd0390912c2b2702d00a093aacface1d0db711bb052cf712c6c3ad4af0fc429d24a49aa574439d370f1304d79cb46c0a80f17d76e11daba001fdd17ebe9d6e647e191bbfb9e24ae72a37ab5b91114e74ab944cc13894d81ec7411a6018a590960c6f0c9e663b818f2af0e468e22b5a11c1583753d3be769e53e022edb486fe7cc4d26a9754d30d62bcdaa0a69401916c58a5d9359736a0d37689f1a278d01eae1d16ca133d267815975536bf41f11eccec14493063f6a7f5784900ca0e7fec43f4bb466a2c835203c40a8f2fe24e7644473eb7220919a592fc25fdf6d17668758a ...
周报2025年3月8日
854efd16088ec0a2949395f17c9b60f9618ed3796518aea5e690afa16559fc1f518ab69ca7808b7649ca5c3ee41ebab93fe6cbbb9aaf4e4fd2361ce0244f7fa003353c6b5ec5997f72f3f9b478896a3ca1297a5ee985f4807f028cf38fb7c1bc55d91146c73513415935bbc2e99c38a94cdc87e714499b06f1b4be63b2fb33a494f48fdc81272949e28a725f2ebf9caa3197046d39a08d7873950565737ff14e9fe635bf426811306c7289696f4a25393986fab91c9cd6fcd611fc9669d6e2320e1f696e1f6e69ecc6d66775b8f99973eb8b42ad06e773fcfb1a40f8209f2957df1961eeea7c38dae7ab979682f97ccd5a9ee0a5ed2088d23 ...
Self-supervised pre-training for large-scale crop mapping using Sentinel-2 time series
854efd16088ec0a2949395f17c9b60f9386aedb0ff5a819cd0390912c2b2702dbca3b60c1de60625835f90b8e297dcaf50f0d6545a351022f83868dcb1452460881f4776c5b3a9d59ac6c00b79914e262186a0b32e67d694fb57cfb13c441faae35baad131d80e753ad2819ea68df48e9e5347a099379b71ccb70635b3934b68b94237e702ccd529499af1b1b3d237ceb182dc0f6636b3025e8c256b156f456824963b13800c900ae888c3dac0d43c29aa632c3027a8bf1aa284d6d4caa1ce63b912006e2d4663fdafd408d07ed934e57a9667e43ac9f8cddb6c7ba02dfe5c2680f2b7e23c3a535e6e71d392701d00335b2691b48cd7f74e3 ...
Untitled
<Boosting crop classification by hierarchically fusing satellite,
rotational, and contextual data>
（提高国家之间的可转移性)
在数据集上进行预训练可提高国家之间的可转移性，允许进行跨域和标签预测，并且在从法国到荷兰的少样本设置（即当领域根据显著新标签发生变化时）下性能具有性能鲁棒性。
提出的方法通过使深度学习方法能够使用经常被忽视的地块时空上下文而优于可比方法
卫星图像时间序列（SITS）数据集在过去十年中取得了重大发展（Selea，2023），为采用数据密集型方法铺平了道路。
引入了一种新颖的多模态建模方法，该方法通过使用作物轮作（CR，Crop
Rotation）
将地球观测数据与历史种植实践相结合。
这种创新方案能够在测试年份没有地面真值数据的情况下预测作物类型。
以加利福尼亚为研究区域，Zhong
等人（2019）研究了使用一维卷积人工神经网络（ANN）（Conv1D）和长短期记忆网络（LSTM）进行作物类型制图的可行性
在欧洲，Pelletier
等人（2019）是 ...
Spatiotemporal masked pre-training for advancing crop mapping on satellite image time series with limited labels
854efd16088ec0a2949395f17c9b60f9386aedb0ff5a819cd0390912c2b2702da4fa93962893de4047c398e96024285cf30a41b7f88a4850535debcc4856a777b7acb5534e8f6259928a36809d7ee5cfc7268bd161b2f62466679515b7dd38fa73f477a4e25d732b989099b19b17dd694fba27922dcb74bb59fbaec926d366a02d6229f3449032f4fb3acbf76ae8452a55a312dacf95798e2075381f2b996649a0b9986791d9bf83c90c177a7d4557422b405fd3be8839a19d90f8b765a1c637d3078d67e4e504335036d33c5a61567ffae9ed7e651a3de5346cc93fa5be399aacc176815214547cbcf4eee4e379e9ea6b2dc4ea8de5b81dc ...
Demonstration of large area land cover classification with a one dimensional convolutional neural network applied to single pixel temporal metric percentiles
854efd16088ec0a2949395f17c9b60f9386aedb0ff5a819cd0390912c2b2702d65e0e58a9ce20c4cfea600f71dc8eabf620b228441a308891ac228094cbde8646ced1527d1e28291a8a9678504ea6d6b8d3822a3cbb82109f9668bec30a0cbb73e9485e213cacbaff52802a6523ce8d5a9dff01c4d1e17cae860cd00e441f8502f7eaa015d9d3857f0ceb1b55a8ade7529155a3c97fcdc0f0e167e5252aee81819ad607159f03cd540a4ec89149cb09d302ca1ad4ebe64a310b08ccd1f8bdb5ea0ee1fc7403a812d4cbf6086ae0dea8c666cbba1d062e845ea3ec6a1cc43679e871e9d975722db621af22079c41a5e8f9ebb90b75440eb412 ...
BERT Pre-training of Deep Bidirectional Transformers for Language Understanding
854efd16088ec0a2949395f17c9b60f9386aedb0ff5a819cd0390912c2b2702d5eeebe3d1d783814f6e286465f1cfee9bc6db2fba24c9157bd124ca7eeb7c0d4895b43645028c4a6897840f061ec728494bd72b8ea1657668476056793075a31fe0e412d2535c2d398cdec3f4fdd5bd79653d1e7f20553d7481c8e702d58a96653f8f7c49c861db54f70405b16bc4399507234931504ea1e39f1ed35e66e259082769c3cf0473e0eb54f5a45ec68664ca7215433c1832facf6ac9b2483a335d21836e4dfa9a0fcdaa28d97ff5c8ddd5d787709de407e6ea189305f13a4868982607921dea556471199f7b0c317c624d09ca6749871c9ca02e ...
周报2025年3月1日
854efd16088ec0a2949395f17c9b60f9618ed3796518aea5e690afa16559fc1f16ad65adf5f3c65c1a672ae8f617fb2526cbd5837db3678003a2a418fe6d8f32b6162731d53757c776f3a02c3e682cfb1d91b9317715732091fb6876d2be4c1a4bcf157f2bff6ff8db34e1fbfeacf778e8e3ad3c7f069aec35edeec4a8b32b5aba7407271d6278008be119b4a126a89af9487ea0aacae21ecff5ca84b9526861967277ad3aa021a5621c506353cef98503ac76fbe708d919b5fcbb73c3d674bccfee0e6863f13e8d9bbe95e92fc9a9c1e0466666a91335f7a41b28e8a51ab14678e7fec87bef750da9f6833c488cfcee2c81f0b627be14b28 ...
Linearly interpolating missing values in time series helps little for land cover classification using recurrent or attention networks
854efd16088ec0a2949395f17c9b60f9386aedb0ff5a819cd0390912c2b2702d5ca5cefc5dff0f42d32f851611d4fac5681b05b249df8eb53ca94c0e9dd96a324fb90dea4407016b25586eb2cc633dc9de1242daf7e6823ec509b5523bc4e9d41e320db048f4a3d4ed681c2f902662f1eac346914c3c21cfa07909bb12eab41fdbb6795c92070efacdcf880e3fcf4b580fd5a4472f76af66b62432cf0612eaa42a2488d19123ffb4de181addb9743cf6c4aa523dfbebb652c726ce5dc8cb38b2c990dff7b2bae0a635da87fc2de36a40b26c4d029008eff07222dd400de90d8eef5f9766e9edde8ce8c4cb75e6296221f62853bf264044ef2 ...
End-to-end learned early classification of time series for in-season crop type mapping
854efd16088ec0a2949395f17c9b60f9386aedb0ff5a819cd0390912c2b2702dbc14a33ebe00ea7297e4c5cdcf59d739a65a43e12254340bf9c8070bf98b1a4898e4adaa117fa0eea8108f3b03fe56dcff4bf8ed57fd8a2d2ca19bca35bd20c5e035f21329588d503ef52924454b20ecacaa726c090a06d23e4407e90b3c30b5f969b0f3d43efa9f6cad5bbe0f1ef446562653f074b17e6e5c01fef20582ff08723712304c557d13021ce27dce1c6caf72a05d0c35f46bba11d5ceb5e3c3726e604ea0ef28ee8b70454686ef2a10a936eba52feb03e2e20d7f7f8e7020e433fec8f5027fc1ef3b661d458bcdf0a220f61576d54a4d888d2f3 ...