# Font Cross-Reconstruction on Chinese font dataset (cross-reconstruction protocol) Source: [Font Cross-Reconstruction on Chinese font dataset (cross-reconstruction protocol)]() Collection index: [llms.txt]() Data updated: 2026-04-09T23:28:19.984Z Evaluation of models performing Font Cross-Reconstruction on the Chinese font dataset using the cross-reconstruction protocol, measured by IoU, L1, RE, Cont. Acc., and Align. Acc. 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 | Font Cross-Reconstruction | | Dataset (including any stated split) | Chinese font dataset (cross-reconstruction protocol) | | Metric | IoU | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 5 of 5 | | 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 - [IoU](): Higher is better; unit: PERCENT. - [L1 Loss](): Lower is better; unit: SCALAR. - [RE](): Lower is better; unit: SCALAR. - [Continuous Accuracy](): Higher is better; unit: PERCENT. - [Alignment Accuracy](): Higher is better; unit: PERCENT. ## Recorded results | Method | IoU | Result date | Report paper | | --- | --- | --- | --- | | DeepVecFont-v2 | 41.7 | Not recorded | 69d732787d0172140ec76a82 | | DesigNet w/o Self-Ref. | 39.7 | Not recorded | 69d732787d0172140ec76a82 | | DesigNet | 39.1 | Not recorded | 69d732787d0172140ec76a82 | | DeepVecFont-v2 | 38 | Not recorded | 69d732787d0172140ec76a82 | | DeepVecFont-v2 + Self-Ref. | 37.8 | Not recorded | 69d732787d0172140ec76a82 | ### DeepVecFont-v2 IoU: 41.7. Result ID: 69d8360347f34ce924bff33f. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Shots":"10"}. This record does not assert independent reproduction. ### DesigNet w/o Self-Ref. IoU: 39.7. Result ID: 69d8360347f34ce924bff33d. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Self-Refinement":"false","Encoder layers":"5","Decoder layers":"5","Shots":"1"}. This record does not assert independent reproduction. ### DesigNet IoU: 39.1. Result ID: 69d8360347f34ce924bff33e. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Self-Refinement":"true","Encoder layers":"5","Decoder layers":"5","Shots":"1"}. This record does not assert independent reproduction. ### DeepVecFont-v2 IoU: 38. Result ID: 69d8360347f34ce924bff33b. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Shots":"1"}. This record does not assert independent reproduction. ### DeepVecFont-v2 + Self-Ref. IoU: 37.8. Result ID: 69d8360347f34ce924bff33c. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Self-Refinement":"true","Shots":"1"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.