ICLR 2021 accepted papers on Detection and segmentation
22 of the 860 papers accepted at ICLR 2021 (2.6%) are on Detection and segmentation, compared with 2.0% at ICLR 2020. They fall into 4 topics.
See ICLR 2021 in the interactive map
Submitting to ICLR? Checklist →
Working on something in this area? Describe your idea in scime atlas to see the closest papers by meaning and how crowded the spot has become.
Topics and papers
CNN features · 12 papers
- Deformable DETR: Deformable Transformers for End-to-End Object Detection (1,852 citations)
- SaliencyMix: A Saliency Guided Data Augmentation Strategy for Better Regularization (101 citations)
- Mind the Pad -- CNNs Can Develop Blind Spots (43 citations)
- Attentional Constellation Nets for Few-Shot Learning (42 citations)
- Estimating and Evaluating Regression Predictive Uncertainty in Deep Object Detectors (24 citations)
- Loss Function Discovery for Object Detection via Convergence-Simulation Driven Search (17 citations)
- Shape or Texture: Understanding Discriminative Features in CNNs (15 citations)
- Topology-Aware Segmentation Using Discrete Morse Theory (14 citations)
- Auto Seg-Loss: Searching Metric Surrogates for Semantic Segmentation (12 citations)
- LiftPool: Bidirectional ConvNet Pooling (9 citations)
- PolarNet: Learning to Optimize Polar Keypoints for Keypoint Based Object Detection (4 citations)
- End-to-End Egospheric Spatial Memory (3 citations)
Open-set and few-shot detection · 6 papers
- What Makes Instance Discrimination Good for Transfer Learning? (82 citations)
- Noise or Signal: The Role of Image Backgrounds in Object Recognition (57 citations)
- VTNet: Visual Transformer Network for Object Goal Navigation (36 citations)
- Unsupervised Object Keypoint Learning using Local Spatial Predictability (7 citations)
- Learning Long-term Visual Dynamics with Region Proposal Interaction Networks (4 citations)
- Generating Furry Cars: Disentangling Object Shape and Appearance across Multiple Domains (3 citations)
