{"url":"/method/deeplabv3","slug":"deeplabv3","name":"DeepLabv3","full_name":"DeepLabv3","full_name_withheld":false,"description_markdown":"**DeepLabv3** is a semantic segmentation architecture that improves upon [DeepLabv2](https://paperswithcode.com/method/deeplabv2) with several modifications. To handle the problem of segmenting objects at multiple scales, modules are designed which employ atrous [convolution](https://paperswithcode.com/method/convolution) in cascade or in parallel to capture multi-scale context by adopting multiple atrous rates. Furthermore, the Atrous [Spatial Pyramid Pooling](https://paperswithcode.com/method/spatial-pyramid-pooling) module from DeepLabv2 augmented with image-level features encoding global context and further boost performance. \r\n\r\nThe changes to the ASSP module are that the authors apply [global average pooling](https://paperswithcode.com/method/global-average-pooling) on the last feature map of the model, feed the resulting image-level features to a 1 × 1 convolution with 256 filters (and [batch normalization](https://paperswithcode.com/method/batch-normalization)), and then bilinearly upsample the feature to the desired spatial dimension. In the\r\nend, the improved [ASPP](https://paperswithcode.com/method/aspp) consists of (a) one 1×1 convolution and three 3 × 3 convolutions with rates = (6, 12, 18) when output stride = 16 (all with 256 filters and batch normalization), and (b) the image-level features.\r\n\r\nAnother interesting difference is that DenseCRF post-processing from DeepLabv2 is no longer needed.","description_state":"present","introduced_year":null,"introduced_by":{"title":null,"paper":null,"first_author":null,"n_authors":0,"url_abs":null,"archive_paper_url":null},"source":{"url":"http://arxiv.org/abs/1706.05587v3","title":"Rethinking Atrous Convolution for Semantic Image Segmentation","url_on_a_paper_host":true},"code_snippet_url":"https://github.com/tensorflow/models/tree/master/research/deeplab","code_snippet_url_on_a_code_host":true,"categories":[{"area":"Computer Vision","area_id":"computer-vision","collection":"Semantic Segmentation Models","url":"/methods/category/semantic-segmentation-models","pwc_aliases":["segmentation-models"]}],"n_papers_tagged":53,"archive_num_papers":null,"papers_newest_first":[{"paper":null,"title":"SAR Object Detection with Self-Supervised Pretraining and Curriculum-Aware Sampling","date":"2025-04-17","arxiv_id":"2504.13310","n_code_links":0,"syntology":null},{"paper":null,"title":"Evaluating and Enhancing Segmentation Model Robustness with Metamorphic Testing","date":"2025-04-03","arxiv_id":"2504.02335","n_code_links":0,"syntology":null},{"paper":null,"title":"DuckSegmentation: A segmentation model based on the AnYue Hemp Duck Dataset","date":"2025-03-27","arxiv_id":"2503.21323","n_code_links":0,"syntology":null},{"paper":"/paper/ai-driven-water-segmentation-with-deep","title":"AI Driven Water Segmentation with deep learning models for Enhanced Flood Monitoring","date":"2025-01-14","arxiv_id":"2501.08266","n_code_links":1,"syntology":null},{"paper":"/paper/mapping-africa-settlements-high-resolution","title":"Mapping Africa Settlements: High Resolution Urban and Rural Map by Deep Learning and Satellite Imagery","date":"2024-11-05","arxiv_id":"2411.02935","n_code_links":1,"syntology":null},{"paper":null,"title":"Automated Surgical Skill Assessment in Endoscopic Pituitary Surgery using Real-time Instrument Tracking on a High-fidelity Bench-top Phantom","date":"2024-09-25","arxiv_id":"2409.17025","n_code_links":0,"syntology":null},{"paper":"/paper/a-novel-open-source-ultrasound-dataset-with","title":"A novel open-source ultrasound dataset with deep learning benchmarks for spinal cord injury localization and anatomical segmentation","date":"2024-09-24","arxiv_id":"2409.16441","n_code_links":1,"syntology":null},{"paper":null,"title":"WaterMAS: Sharpness-Aware Maximization for Neural Network Watermarking","date":"2024-09-05","arxiv_id":"2409.03902","n_code_links":0,"syntology":null},{"paper":null,"title":"Deep Convolutional Neural Networks Meet Variational Shape Compactness Priors for Image Segmentation","date":"2024-05-23","arxiv_id":"2406.19400","n_code_links":0,"syntology":null},{"paper":"/paper/one-model-to-use-them-all-training-a","title":"One model to use them all: Training a segmentation model with complementary datasets","date":"2024-02-29","arxiv_id":"2402.19340","n_code_links":1,"syntology":null},{"paper":null,"title":"CAFCT-Net: A CNN-Transformer Hybrid Network with Contextual and Attentional Feature Fusion for Liver Tumor Segmentation","date":"2024-01-30","arxiv_id":"2401.16886","n_code_links":0,"syntology":null},{"paper":"/paper/mixmobilenet-a-mixed-mobile-network-for-edge","title":"MixMobileNet: A Mixed Mobile Network for Edge Vision Applications","date":"2024-01-26","arxiv_id":null,"n_code_links":1,"syntology":null},{"paper":null,"title":"Scaling Up Quantization-Aware Neural Architecture Search for Efficient Deep Learning on the Edge","date":"2024-01-22","arxiv_id":"2401.12350","n_code_links":0,"syntology":null},{"paper":"/paper/vrem-fl-mobility-aware-computation-scheduling","title":"VREM-FL: Mobility-Aware Computation-Scheduling Co-Design for Vehicular Federated Learning","date":"2023-11-30","arxiv_id":"2311.18741","n_code_links":1,"syntology":null},{"paper":"/paper/dacl10k-benchmark-for-semantic-bridge-damage","title":"dacl10k: Benchmark for Semantic Bridge Damage Segmentation","date":"2023-09-01","arxiv_id":"2309.00460","n_code_links":1,"syntology":null},{"paper":null,"title":"Feature Selection on Sentinel-2 Multi-spectral Imagery for Efficient Tree Cover Estimation","date":"2023-05-31","arxiv_id":"2306.06073","n_code_links":0,"syntology":null},{"paper":null,"title":"Radious: Unveiling the Enigma of Dental Radiology with BEIT Adaptor and Mask2Former in Semantic Segmentation","date":"2023-05-10","arxiv_id":"2305.06236","n_code_links":0,"syntology":null},{"paper":null,"title":"Detection and Classification of Glioblastoma Brain Tumor","date":"2023-04-18","arxiv_id":"2304.09133","n_code_links":0,"syntology":null},{"paper":null,"title":"Framework-agnostic Semantically-aware Global Reasoning for Segmentation","date":"2022-12-06","arxiv_id":"2212.03338","n_code_links":0,"syntology":null},{"paper":"/paper/isim-iterative-self-improved-model-for-weakly","title":"ISIM: Iterative Self-Improved Model for Weakly Supervised Segmentation","date":"2022-11-22","arxiv_id":"2211.12455","n_code_links":1,"syntology":null},{"paper":"/paper/self-adversarial-multi-scale-contrastive","title":"Self-adversarial Multi-scale Contrastive Learning for Semantic Segmentation of Thermal Facial Images","date":"2022-09-21","arxiv_id":"2209.10700","n_code_links":1,"syntology":{"ran":1,"of":3,"unverified":2,"pointer_only":0}},{"paper":null,"title":"Multiclass ASMA vs Targeted PGD Attack in Image Segmentation","date":"2022-08-03","arxiv_id":"2208.01844","n_code_links":0,"syntology":null},{"paper":"/paper/transdeeplab-convolution-free-transformer","title":"TransDeepLab: Convolution-Free Transformer-based DeepLab v3+ for Medical Image Segmentation","date":"2022-08-01","arxiv_id":"2208.00713","n_code_links":1,"syntology":null},{"paper":"/paper/masked-generative-distillation","title":"Masked Generative Distillation","date":"2022-05-03","arxiv_id":"2205.01529","n_code_links":3,"syntology":null},{"paper":"/paper/signature-and-log-signature-for-the-study-of","title":"Signature and Log-signature for the Study of Empirical Distributions Generated with GANs","date":"2022-03-07","arxiv_id":"2203.03226","n_code_links":1,"syntology":null},{"paper":"/paper/muad-multiple-uncertainties-for-autonomous","title":"MUAD: Multiple Uncertainties for Autonomous Driving, a benchmark for multiple uncertainty types and tasks","date":"2022-03-02","arxiv_id":"2203.01437","n_code_links":3,"syntology":null},{"paper":"/paper/loveda-a-remote-sensing-land-cover-dataset","title":"LoveDA: A Remote Sensing Land-Cover Dataset for Domain Adaptive Semantic Segmentation","date":"2021-10-17","arxiv_id":"2110.08733","n_code_links":5,"syntology":{"ran":2,"of":6,"unverified":4,"pointer_only":0}},{"paper":"/paper/divergentnets-medical-image-segmentation-by","title":"DivergentNets: Medical Image Segmentation by Network Ensemble","date":"2021-07-01","arxiv_id":"2107.00283","n_code_links":1,"syntology":null},{"paper":"/paper/encoder-decoder-architectures-for-clinically","title":"Encoder-Decoder Architectures for Clinically Relevant Coronary Artery Segmentation","date":"2021-06-21","arxiv_id":"2106.11447","n_code_links":1,"syntology":null},{"paper":"/paper/reducing-the-feature-divergence-of-rgb-and","title":"Reducing the feature divergence of RGB and near-infrared images using Switchable Normalization","date":"2021-06-06","arxiv_id":"2106.03088","n_code_links":1,"syntology":null}],"papers_shown":30,"tasks":[{"task":"/task/semantic-segmentation","name":"Semantic Segmentation","papers":40},{"task":"/task/segmentation","name":"Segmentation","papers":25},{"task":"/task/image-segmentation","name":"Image Segmentation","papers":14},{"task":"/task/object-detection","name":"Object Detection","papers":11},{"task":"/task/object-detection-1","name":"object-detection","papers":8},{"task":"/task/image-classification","name":"Image Classification","papers":7},{"task":"/task/instance-segmentation","name":"Instance Segmentation","papers":7},{"task":"/task/image-classification","name":"image-classification","papers":5},{"task":"/task/classification-1","name":"Classification","papers":3},{"task":"/task/decoder","name":"Decoder","papers":3},{"task":"/task/architecture-search","name":"Neural Architecture Search","papers":3},{"task":"/task/object","name":"Object","papers":3},{"task":"/task/anatomy","name":"Anatomy","papers":2},{"task":"/task/anomaly-detection","name":"Anomaly Detection","papers":2},{"task":"/task/autonomous-driving","name":"Autonomous Driving","papers":2},{"task":"/task/depth-estimation","name":"Depth Estimation","papers":2},{"task":"/task/domain-adaptation","name":"Domain Adaptation","papers":2},{"task":"/task/image-augmentation","name":"Image Augmentation","papers":2},{"task":"/task/knowledge-distillation","name":"Knowledge Distillation","papers":2},{"task":"/task/medical-image-analysis","name":"Medical Image Analysis","papers":2}],"tasks_shown":20,"n_tasks":97,"usage_by_year":[{"year":"2017","papers":1},{"year":"2018","papers":3},{"year":"2019","papers":5},{"year":"2020","papers":7},{"year":"2021","papers":11},{"year":"2022","papers":8},{"year":"2023","papers":5},{"year":"2024","papers":9},{"year":"2025","papers":4}],"row_source":"embedded","archive":{"source":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","archive_url":"https://paperswithcode.com/method/deeplabv3"},"syntology_read_at":"2026-09-24T18:15:14+00:00"}