# End-to-End Forecasting on Argoverse 2 (val) Source: [End-to-End Forecasting on Argoverse 2 (val)]() Collection index: [llms.txt]() Data updated: 2026-03-11T08:45:50.155Z Evaluation of end-to-end forecasting models on the Argoverse 2 validation set, assessed using Collision Detection Score (CDS) and Average Precision (AP) metrics across various prediction horizons. 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 | End-to-End Forecasting | | Dataset (including any stated split) | Argoverse 2 (val) | | Metric | CDS (1-20) | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 6 of 6 | | 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 - [CDS (1-20)](): Higher is better; unit: PERCENT. - [CDS (21-26)](): Higher is better; unit: PERCENT. - [CDS (All)](): Higher is better; unit: PERCENT. - [AP Static (1-20)](): Higher is better; unit: PERCENT. - [AP Static (21-26)](): Higher is better; unit: PERCENT. - [AP Static (All)](): Higher is better; unit: PERCENT. - [AP Linear (1-20)](): Higher is better; unit: PERCENT. - [AP Linear (21-26)](): Higher is better; unit: PERCENT. - [AP Linear (All)](): Higher is better; unit: PERCENT. - [AP Non-Linear (1-20)](): Higher is better; unit: PERCENT. - [AP Non-Linear (21-26)](): Higher is better; unit: PERCENT. - [AP Non-Linear (All)](): Higher is better; unit: PERCENT. - [mAP (1-20)](): Higher is better; unit: PERCENT. - [mAP (21-26)](): Higher is better; unit: PERCENT. - [mAP (All)](): Higher is better; unit: PERCENT. ## Recorded results | Method | CDS (1-20) | Result date | Report paper | | --- | --- | --- | --- | | Joint Training | 31.21 | Not recorded | 69b0f7292d603e6de3e45f31 | | OMEN | 29.19 | Not recorded | 69b0f7292d603e6de3e45f31 | | Pseudo-Labeling | 29.01 | Not recorded | 69b0f7292d603e6de3e45f31 | | Pseudo-Labeling | 28.54 | Not recorded | 69b0f7292d603e6de3e45f31 | | CL-DETR | 28.43 | Not recorded | 69b0f7292d603e6de3e45f31 | | Forgetting | 0 | Not recorded | 69b0f7292d603e6de3e45f31 | ### Joint Training CDS (1-20): 31.21. Result ID: 69b12bb96dbb7075b95ba877. [Reporting paper]() Recorded evaluation configuration: {"Backbone":"VoVNet-99","Evaluation range":"50m"}. This record does not assert independent reproduction. ### OMEN CDS (1-20): 29.19. Result ID: 69b12bb96dbb7075b95ba87c. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"VoVNet-99","Evaluation range":"50m"}. This record does not assert independent reproduction. ### Pseudo-Labeling CDS (1-20): 29.01. Result ID: 69b12bb96dbb7075b95ba879. [Reporting paper]() Recorded evaluation configuration: {"Backbone":"VoVNet-99","Evaluation range":"50m"}. This record does not assert independent reproduction. ### Pseudo-Labeling CDS (1-20): 28.54. Result ID: 69b12bb96dbb7075b95ba87a. [Reporting paper]() Recorded evaluation configuration: {"Backbone":"VoVNet-99","Evaluation range":"50m","Pseudo-labeling from future detections":"true"}. This record does not assert independent reproduction. ### CL-DETR CDS (1-20): 28.43. Result ID: 69b12bb96dbb7075b95ba87b. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"VoVNet-99","Evaluation range":"50m","Pseudo-labeling from future detections":"true"}. This record does not assert independent reproduction. ### Forgetting CDS (1-20): 0. Result ID: 69b12bb96dbb7075b95ba878. [Reporting paper]() Recorded evaluation configuration: {"Backbone":"VoVNet-99","Evaluation range":"50m"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.