FEELVOS: Fast End-to-End Embedding Learning for Video Object Segmentation

CVPR 2019 Paul VoigtlaenderYuning ChaiFlorian SchroffHartwig AdamBastian LeibeLiang-Chieh Chen

Many of the recent successful methods for video object segmentation (VOS) are overly complicated, heavily rely on fine-tuning on the first frame, and/or are slow, and are hence of limited practical use. In this work, we propose FEELVOS as a simple and fast method which does not rely on fine-tuning... (read more)

PDF Abstract
TASK DATASET MODEL METRIC NAME METRIC VALUE GLOBAL RANK RESULT BENCHMARK
Semi-Supervised Video Object Segmentation DAVIS 2016 FEELVOS Jaccard (Mean) 81.1 # 16
Jaccard (Recall) 90.5 # 17
Jaccard (Decay) 13.7 # 5
F-measure (Mean) 82.2 # 11
F-measure (Recall) 86.6 # 18
F-measure (Decay) 14.1 # 20
J&F 81.65 # 14
Semi-Supervised Video Object Segmentation DAVIS 2017 (test-dev) FEELVOS J&F 57.8 # 6
Jaccard (Mean) 55.1 # 6
Jaccard (Recall) 62.6 # 6
Jaccard (Decay) 29.8 # 14
F-measure (Mean) 60.4 # 7
F-measure (Recall) 68.5 # 7
F-measure (Decay) 33.5 # 14
Semi-Supervised Video Object Segmentation DAVIS 2017 (val) FEELVOS Jaccard (Mean) 69.1 # 5
Jaccard (Recall) 79.1 # 4
Jaccard (Decay) 17.5 # 7
F-measure (Mean) 74.0 # 6
F-measure (Recall) 83.8 # 4
F-measure (Decay) 20.1 # 10
J&F 71.55 # 5
Semi-Supervised Video Object Segmentation YouTube FEELVOS mIoU 0.821 # 1

Methods used in the Paper


METHOD TYPE
🤖 No Methods Found Help the community by adding them if they're not listed; e.g. Deep Residual Learning for Image Recognition uses ResNet