# Object Segmentation on PartImageNet (val) Source: [Object Segmentation on PartImageNet (val)]() Collection index: [llms.txt]() Data updated: 2026-02-26T06:10:23.100Z Evaluation of Object Segmentation performance on the validation split of the PartImageNet dataset, reporting metrics such as mIoU and mACC. A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage. | Scope | Recorded value | | --- | --- | | Task | Object Segmentation | | Dataset (including any stated split) | PartImageNet (val) | | Metric | mIoU | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 9 of 9 | | Coverage | All results currently returned for this benchmark | Evaluation settings may differ between results. The original papers and recorded configurations are needed to establish comparability. A shared metric name alone does not establish an identical protocol. ## Metrics - [mIoU](): Higher is better; unit: PERCENT. - [mACC](): Higher is better; unit: PERCENT. ## Recorded results | Method | mIoU | Result date | Report paper | | --- | --- | --- | --- | | Compositor | 78.98 | Not recorded | 68fa715a1e209131cea50c93 | | MaskFormer | 77.92 | Not recorded | 68fa715a1e209131cea50c93 | | MaskFormer-Dual | 77.24 | Not recorded | 68fa715a1e209131cea50c93 | | SegFormer | 74.55 | Not recorded | 68fa715a1e209131cea50c93 | | Compositor | 71.78 | Not recorded | 68fa715a1e209131cea50c93 | | MaskFormer-Dual | 70.44 | Not recorded | 68fa715a1e209131cea50c93 | | MaskFormer | 70.21 | Not recorded | 68fa715a1e209131cea50c93 | | Deeplab v3+ | 69.82 | Not recorded | 68fa715a1e209131cea50c93 | | Deeplab v3+ | 68.38 | Not recorded | 68fa715a1e209131cea50c93 | ### Compositor mIoU: 78.98. Result ID: 6991acef3a6c31484909d1fc. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"backbone":"Swin-T","params":"51M","training strategy":"Models on parts and objects are trained jointly"}. This record does not assert independent reproduction. ### MaskFormer mIoU: 77.92. Result ID: 6991acef3a6c31484909d1fa. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"backbone":"Swin-T","params":"92M (46M×2)","training strategy":"Models on parts and objects are trained separately"}. This record does not assert independent reproduction. ### MaskFormer-Dual mIoU: 77.24. Result ID: 6991acef3a6c31484909d1fb. [Reporting paper]() Recorded evaluation configuration: {"backbone":"Swin-T","params":"51M","training strategy":"Models on parts and objects are trained jointly"}. This record does not assert independent reproduction. ### SegFormer mIoU: 74.55. Result ID: 6991acef3a6c31484909d1f9. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"backbone":"MiT-B2","params":"48M (24M×2)","training strategy":"Models on parts and objects are trained separately"}. This record does not assert independent reproduction. ### Compositor mIoU: 71.78. Result ID: 6991acef3a6c31484909d1f8. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"backbone":"ResNet-50","params":"50M","training strategy":"Models on parts and objects are trained jointly"}. This record does not assert independent reproduction. ### MaskFormer-Dual mIoU: 70.44. Result ID: 6991acef3a6c31484909d1f7. [Reporting paper]() Recorded evaluation configuration: {"backbone":"ResNet-50","params":"50M","training strategy":"Models on parts and objects are trained jointly"}. This record does not assert independent reproduction. ### MaskFormer mIoU: 70.21. Result ID: 6991acef3a6c31484909d1f6. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"backbone":"ResNet-50","params":"90M (45M×2)","training strategy":"Models on parts and objects are trained separately"}. This record does not assert independent reproduction. ### Deeplab v3+ mIoU: 69.82. Result ID: 6991acef3a6c31484909d1f5. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"backbone":"ResNet-50","params":"42M","training strategy":"Models on objects are trained with parts as deep supervision"}. This record does not assert independent reproduction. ### Deeplab v3+ mIoU: 68.38. Result ID: 6991acef3a6c31484909d1f4. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"backbone":"ResNet-50","params":"84M (42M×2)","training strategy":"Models on parts and objects are trained separately"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.