search

JialeCao001 / SipMask

SipMask: Spatial Information Preservation for Fast Image and Video Instance Segmentation (ECCV2020)
https://arxiv.org/pdf/2007.14772.pdf
MIT License
334 stars 54 forks source link

SipMask_VIS:when I using the pertrained model to test the youtube dataset, the result of model is None? #45

Open sally1913105 opened 3 years ago

sally1913105 commented 3 years ago

image Maybe the scores were too low and filtered, but I tried the YouTubeVIS valid datasets and the model detected nothing

JialeCao001 commented 3 years ago

@sally1913105 Can you provide more details.