Browse › Robots › CARLA longest6 › CARLA
CARLA Benchmark (CARLA longest6)
longest6 is an evaluation benchmark for sensorimotor autonomous driving methods using the CARLA 0.9.10.1 simulator. It consists of 36 long routes in the publicly available Town 01-06 which, are populated with the maximum traffic density. The benchmark tests level 4 driving capabilities, methods are therefore allowed to train with data from the evaluation towns. Evaluation metrics follow the standard metrics from the CARLA leaderboard 1.0, except that stop sign infractions are not considered.
The archive carries no text for this table; the description above is the archive's text for the task CARLA longest6. archive 2025-07-28
Over time archive 2025-07-28
The chart needs JavaScript; the table below carries every value.
Direction inferred from the metric name, not from the archive: Driving Score (higher is better), Infraction Score (higher is better). Not inferred (points only, no best-so-far line): Route Completion. Points are placed at the row's paper date; 19 of 21 rows carry one.
Results archive 2025-07-28
Archive rows end at the archive snapshot, 2025-07-28: no result published after that date is in this table. Rank is the archive's row order at that snapshot; not re-ranked here. Metric values are the archive's strings. Column headers sort the table in your browser; each row keeps its archive rank.
| Paper | Code | Ran Syntology | Report | |||||||
|---|---|---|---|---|---|---|---|---|---|---|
| 1 | DiffE2E | 83 | 96 | 0.86 | – | – | – | not matched | report | |
| 2 | TransFuser++ WP (TF++WP) | 73 | 97 | 0.56 | – | Paper | Code | 2023 | 13 of 17 ran · 4 unverified | report |
| 3 | DriveAdapter+TCP | 71 | 88 | 0.85 | – | Paper | Code | 2023 | 6 of 7 ran · 1 unverified | report |
| 4 | DriveGPT4-V2 | 70 | 91 | 0.77 | – | – | – | not matched | report | |
| 5 | EnDfuser + safety-rule | 70 | 90 | 0.76 | – | Paper | – | 2025 | no code linked | report |
| 6 | TransFuser++ (TF++) | 69 | 94 | 0.72 | – | Paper | Code | 2023 | 13 of 17 ran · 4 unverified | report |
| 7 | ThinkTwice | 67 | 77 | 0.84 | – | Paper | Code | 2023 | 4 of 7 ran · 3 unverified | report |
| 8 | DriveAdapter | 59 | 82 | 0.68 | – | Paper | Code | 2023 | 6 of 7 ran · 1 unverified | report |
| 9 | Perception PlanT | 58 | 88 | 0.65 | – | Paper | Code | 2022 | 2 of 2 ran · 0 unverified | report |
| 10 | Learning from all Vehicle v2 (LAV v2) | 58 | 83 | 0.68 | – | Paper | Code | 2022 | 0 of 5 ran · 5 unverified | report |
| 11 | Coaching a Teachable Student (CaT) | 58 | 79 | 0.77 | – | Paper | Code | 2023 | 0 of 4 ran · 4 unverified | report |
| 12 | InteractionNet | 51 | 87 | 0.60 | – | Paper | Code | 2023 | linked, not harvested | report |
| 13 | Trajectory-guided Control Prediction (TCP) | 48 | 72 | 0.65 | – | Paper | Code | 2022 | linked, not harvested | report |
| 14 | TransFuser (TF) | 47 | 93 | 0.50 | – | Paper | Code | 2022 | 1 of 5 ran · 4 unverified | report |
| 15 | Interfuser | 47 | 74 | 0.63 | – | Paper | Code | 2022 | 1 of 1 ran · 0 unverified | report |
| 16 | Latent TransFuser (LTF) | 37 | 95 | 0.38 | – | Paper | Code | 2022 | 1 of 5 ran · 4 unverified | report |
| 17 | Learning from all Vehicles v1 (LAV v1) | 33 | 70 | 0.51 | – | Paper | Code | 2022 | 0 of 5 ran · 5 unverified | report |
| 18 | Geometric Fusion (GF) | 27 | 91 | 0.30 | – | Paper | Code | 2022 | 1 of 5 ran · 4 unverified | report |
| 19 | Neural Attention Fields (NEAT) | 24 | 62 | 0.71 | – | Paper | Code | 2021 | 0 of 2 ran · 2 unverified | report |
| 20 | Late Fusion (LF) | 22 | 83 | 0.27 | – | Paper | Code | 2022 | 1 of 5 ran · 4 unverified | report |
| 21 | World on Rails (WOR) | 21 | 48 | 0.56 | – | Paper | Code | 2021 | 2 of 6 ran · 4 unverified | report |
All 21 rows shown. 19 link to a paper page on this site; 0 are marked as using additional training data in the archive. No GitHub stars are tracked; "Code" is the first repository the archive lists for the row. The archive carries no row tags, review links or community-submitted rows for this table; none are shown. archive 2025-07-28
Syntology Ran reads "N of M ran · U unverified": of the M code samples Syntology harvested from repositories linked to that row's paper (joined by arXiv id), N executed on a synthesized input and the other U = M−N are unverified (harvested, no recorded run). It counts code from repositories linked to that row's paper, not this result: the row's number was not reproduced and nothing here is a correctness claim. The other cell texts mean no graph line for the row: "linked, not harvested" (the archive links code, Syntology has not harvested it), "no code linked" (no code link in the archive), "not matched" (the row's paper URL matched no paper on this site). 16 rows have a graph line, from 10 distinct papers; 12 rows (7 papers) have at least one sample that ran. Counting each paper once: Syntology ran 29 of 56 samples; 27 unverified. Separately, 0 of those 56 are pointer-only (licence): the site points at that code rather than redistributing it, a licence property recorded for ran and unverified samples alike; each cell's tooltip carries the row's own pointer-only count. Read from the graph 2026-09-24. Per-sample status is on the paper page.
Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections