# Described Object Detection on D3 XL Source: [Described Object Detection on D3 XL]() Collection index: [llms.txt]() Data updated: 2026-03-24T16:28:23.467Z Evaluation of Described Object Detection performance on the D3 XL dataset, reporting mAP scores. 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 | Described Object Detection | | Dataset (including any stated split) | D3 XL | | Metric | mAP | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 14 of 14 | | 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 - [mAP](): Higher is better; unit: PERCENT. ## Recorded results | Method | mAP | Result date | Report paper | | --- | --- | --- | --- | | APE-Ti + NEGTOME | 25.4 | Not recorded | 68fa7c331e209131cea8cfb7 | | APE-Ti | 21.4 | Not recorded | 68fa7c331e209131cea8cfb7 | | G-DINO-B + NEGTOME | 21.3 | Not recorded | 68fa7c331e209131cea8cfb7 | | OFA-DOD | 18.4 | Not recorded | 68fa7c331e209131cea8cfb7 | | Qwen-2.5-VL-3B + NEGTOME | 17.8 | Not recorded | 68fa7c331e209131cea8cfb7 | | G-DINO-B | 16.5 | Not recorded | 68fa7c331e209131cea8cfb7 | | Qwen-2.5-VL-3B | 16 | Not recorded | 68fa7c331e209131cea8cfb7 | | FIBER-B + GEN | 14.2 | Not recorded | 68fa7c331e209131cea8cfb7 | | FIBER-B | 13.1 | Not recorded | 68fa7c331e209131cea8cfb7 | | GLIP-T + GEN | 11.5 | Not recorded | 68fa7c331e209131cea8cfb7 | | GLIP-T | 10.6 | Not recorded | 68fa7c331e209131cea8cfb7 | | OWL-ViT-L | 5.3 | Not recorded | 68fa7c331e209131cea8cfb7 | | SPHINX-7B | 3.1 | Not recorded | 68fa7c331e209131cea8cfb7 | | OFA-L | 2.1 | Not recorded | 68fa7c331e209131cea8cfb7 | ### APE-Ti + NEGTOME mAP: 25.4. Result ID: 69c2bb9347f34ce924ac3279. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"ViT-Ti","Text Encoder":"CLIP","Detection Head":"DETA"}. This record does not assert independent reproduction. ### APE-Ti mAP: 21.4. Result ID: 69c2bb9347f34ce924ac3278. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"ViT-Ti","Text Encoder":"CLIP","Detection Head":"DETA"}. This record does not assert independent reproduction. ### G-DINO-B + NEGTOME mAP: 21.3. Result ID: 69c2bb9347f34ce924ac3277. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Swin-B","Text Encoder":"BERT","Detection Head":"DINO"}. This record does not assert independent reproduction. ### OFA-DOD mAP: 18.4. Result ID: 69c2bb9347f34ce924ac3271. [Reporting paper]() Recorded evaluation configuration: {"Backbone":"ResNet-101+ViT","Text Encoder":"BART","Detection Head":"Seq2Seq"}. This record does not assert independent reproduction. ### Qwen-2.5-VL-3B + NEGTOME mAP: 17.8. Result ID: 69c2bb9347f34ce924ac327b. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"ViT-H","Text Encoder":"Qwen-2.5","Detection Head":"-"}. This record does not assert independent reproduction. ### G-DINO-B mAP: 16.5. Result ID: 69c2bb9347f34ce924ac3276. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"Swin-B","Text Encoder":"BERT","Detection Head":"DINO"}. This record does not assert independent reproduction. ### Qwen-2.5-VL-3B mAP: 16. Result ID: 69c2bb9347f34ce924ac327a. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"ViT-H","Text Encoder":"Qwen-2.5","Detection Head":"-"}. This record does not assert independent reproduction. ### FIBER-B + GEN mAP: 14.2. Result ID: 69c2bb9347f34ce924ac3275. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Swin-B","Text Encoder":"RoBERTa-B","Detection Head":"DyHead"}. This record does not assert independent reproduction. ### FIBER-B mAP: 13.1. Result ID: 69c2bb9347f34ce924ac3274. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"Swin-B","Text Encoder":"RoBERTa-B","Detection Head":"DyHead"}. This record does not assert independent reproduction. ### GLIP-T + GEN mAP: 11.5. Result ID: 69c2bb9347f34ce924ac3273. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Swin-T","Text Encoder":"BERT","Detection Head":"DyHead"}. This record does not assert independent reproduction. ### GLIP-T mAP: 10.6. Result ID: 69c2bb9347f34ce924ac3272. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"Swin-T","Text Encoder":"BERT","Detection Head":"DyHead"}. This record does not assert independent reproduction. ### OWL-ViT-L mAP: 5.3. Result ID: 69c2bb9347f34ce924ac326f. [Reporting paper]() Recorded evaluation configuration: {"Backbone":"ViT-L","Text Encoder":"CLIP","Detection Head":"OWL-ViT"}. This record does not assert independent reproduction. ### SPHINX-7B mAP: 3.1. Result ID: 69c2bb9347f34ce924ac3270. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"CLIP,DINO-v2, Q-Former","Text Encoder":"LLaMA-2","Detection Head":"-"}. This record does not assert independent reproduction. ### OFA-L mAP: 2.1. Result ID: 69c2bb9347f34ce924ac326e. [Reporting paper]() Recorded evaluation configuration: {"Backbone":"ResNet-101+ViT","Text Encoder":"BART","Detection Head":"Seq2Seq"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.