Repository Issue Activity (beta)

#108 0.40.0: visible_qwen detector false-positives on a clean repo fixture, and `--mark auto` inpaints clean pixels

#105 HEIF sources lose their EXIF orientation: pillow-heif blanks the tag on open

#106 A TIFF source comes out turned AND tagged, so a viewer turns it twice

#98 Every command that re-encodes drops the ICC profile and the orientation tag

#103 `--cpu-offload` never reaches the global stack of `chroma-zimage` or `sdxl-zimage`, and the docs say it does

#100 A 16-bit PNG crashes `erase`, quietly corrupts the LaMa fill, and comes out of the metadata strip at 8 bits

#99 `--humanize` runs before `adaptive_polish` and, from about 6 up, quietly turns it into a no-op

#88 Please integrate Flux 2

#84 出现文字畸变问题

#79 Imagem Damoclis

#16 Add support for removing watermarks from videos

#66 Doesn't remove "AI" from attributes of the file

#46 as if it’s not working

#51 Dose NOT REMOVE SynthID remove-ai-watermarks.

#59 Tencent / 腾讯 support

#49 how to work on face quality?

#54 即梦基础会员，4.0，3:4，2K图，左上角AI生成检测不出

#48 It works! But results are blurry for images containing Chinese text

#36 The watermark detect is not working?

#47 SynthID not removed from ChatGPT image

#37 Add visible watermark removal for Samsung Galaxy AI ("Contenuti generati con la AI")

#35 GPT Image 2 Text Preservation

#41 Unable to obtain normal images after removing invisible watermarks

#14 请问应该怎么设置参数，才可以和官网的“一体式全流程”处理效果一样

#13 Hope to support the image of Doubao