Pix2seq: A Language Modeling Framework for Object Detection

Benchmark Model Rank Results
object-detection-on-coco-minivalPix2seq (ViT-L)#81box AP: 50.0
object-detection-on-coco-minivalPix2seq (R50-C4)#97box AP: 47.3
object-detection-on-coco-minivalPix2seq (ViT-B)#98box AP: 47.1
object-detection-on-coco-minivalPix2seq (R101-DC5)#118box AP: 45.0AP50: 63.2AP75: 48.6APS: 28.2APM: 48.9APL: 60.4
object-detection-on-coco-minivalPix2seq (R50-DC5 )#143box AP: 43.2AP50: 61.0AP75: 46.1APS: 26.6APM: 47APL: 58.6
object-detection-on-coco-minivalPix2seq (R50)#152box AP: 42.6