Explore
Everything I have written or built, as one map. An item hangs from what it builds on. Hover an item and watch two things happen: first what is nearest to it lights up, ring by ring, each with its distance; then a dot walks the reading order through it, numbering every item on the way. Everything the two searches do not reach fades away until the next hover. Click to open it. Have a question instead? Ask .
Tap an item once to light up its connections, tap it again to open it. Hover an item to light up its connections, click to open it. The map is wider than the screen: swipe it sideways.
2 links away, nearest first5 place in the reading order
Tap an item on the map Hover an item on the map and its connections are listed here.
Series, generated from the map
A series is a chain of “builds on” links. The map finds the longest chains; you can also ask for the chain between two items, or everything under one tag in reading order.
From — Adaptive traffic lights ALBERT vs BERT Anchor search Assembled CNN Attention before transformers Attention meets pixels Background subtraction Benchmark datasets BoolGAN Camera calibration Camera pose Camera resectioning CBAM and BAM CDAC CenterNet CLIP Convolutions strike back CTAugment DarkCovidNet DeiT DetectoRS DETR DETR (VisionWizard) DIoU loss Domain adaptation + SSL Dynamic convolution EfficientDet Epipolar geometry Face spoofing detection Fawkes Focal loss FreeMatch Gender from tweets GPT · Part 1 In defense of triplet loss INT8 quantization Lens distortion LISA · instance shadows MAE mAP calculator MixConv Multi-radius Deep SVDD Network pruning Night-time illumination Object detection Object segmentation Object tracking PAMTRI Person re-identification Pinhole camera Points at infinity Query, key, value RegNet Relative pose RT-DETR RT-DETR-World RT-DETRv2 RT-DETRv3 RT-DETRv4 Saliency maps SANet style transfer SCAN Site-specific training SNIPER Speech recognition in 5 steps SpinalNet SpineNet Stereo depth Stereo point cloud Swin Synthetic bounding boxes Text2SQL · Baselines Text2SQL · Datasets Text2SQL · Final insights Text2SQL · Introduction Text2SQL · State of the art The Transformer Thread library Trajectory anomaly TransCoder Triangulation Triplet loss Vehicle counting UI Video semantic segmentation Vision transformers today ViT VQA on 360° images Weakly supervised detection YOLOv4 · Bag of freebies YOLOv4 · Bag of specials YOLOv4 · Final verdict YOLOv4 · Introduction YOLOv4 · Proposed workflow
to — Adaptive traffic lights ALBERT vs BERT Anchor search Assembled CNN Attention before transformers Attention meets pixels Background subtraction Benchmark datasets BoolGAN Camera calibration Camera pose Camera resectioning CBAM and BAM CDAC CenterNet CLIP Convolutions strike back CTAugment DarkCovidNet DeiT DetectoRS DETR DETR (VisionWizard) DIoU loss Domain adaptation + SSL Dynamic convolution EfficientDet Epipolar geometry Face spoofing detection Fawkes Focal loss FreeMatch Gender from tweets GPT · Part 1 In defense of triplet loss INT8 quantization Lens distortion LISA · instance shadows MAE mAP calculator MixConv Multi-radius Deep SVDD Network pruning Night-time illumination Object detection Object segmentation Object tracking PAMTRI Person re-identification Pinhole camera Points at infinity Query, key, value RegNet Relative pose RT-DETR RT-DETR-World RT-DETRv2 RT-DETRv3 RT-DETRv4 Saliency maps SANet style transfer SCAN Site-specific training SNIPER Speech recognition in 5 steps SpinalNet SpineNet Stereo depth Stereo point cloud Swin Synthetic bounding boxes Text2SQL · Baselines Text2SQL · Datasets Text2SQL · Final insights Text2SQL · Introduction Text2SQL · State of the art The Transformer Thread library Trajectory anomaly TransCoder Triangulation Triplet loss Vehicle counting UI Video semantic segmentation Vision transformers today ViT VQA on 360° images Weakly supervised detection YOLOv4 · Bag of freebies YOLOv4 · Bag of specials YOLOv4 · Final verdict YOLOv4 · Introduction YOLOv4 · Proposed workflow
Chain between
or tag — Object detection Anchor-free detection Segmentation Image classification Keypoints and pose Re-identification Tracking Anomaly detection Traffic analytics Saliency Natural language Evaluation and metrics Video Faces Generative models Language models Vision and language Speech and audio Medical imaging Image enhancement Attention modules Self-attention Transformers Vision transformers Positional encoding Convolutions Backbone design Feature pyramids and necks Loss functions Class imbalance Non-maximum suppression Set prediction Activation functions Camera models Projective geometry Camera calibration Two-view geometry 3-D reconstruction Stereo vision Self-supervised learning Semi-supervised learning Domain adaptation Knowledge distillation Data augmentation Training recipes Contrastive learning Pretraining at scale Weak supervision Datasets and labelling Efficient attention Efficient networks Quantization Pruning Edge deployment Systems programming Developer tools
In reading order