AI Paper Atlas

Research map / Benchmarks / COIN T=3

COIN T=3 leaderboard

12 papers at ICML, NeurIPS, ICLR, CVPR, AAAI, ICCV and ECCV, and the journals TPAMI, IJCV and JMLR report results on COIN T=3. Best success rate: 33.99 (ViterbiPlanNet, CVPR 2026).

Numbers are as each paper reports them, read by AI from its results tables and not checked by hand. Papers can differ in setting (training data, model size, extra data); the topic column shows what each paper is about. Open the paper before you compare or quote a number.

COIN T=3 · success rate

12 papers, each with its best reported number for its own method; higher is better.

#success rate ↑MethodPaperTopic
133.99ViterbiPlanNetViterbiPlanNet: Injecting Procedural Knowledge via Differentiable Viterbi for Planning in Instructional Videos
CVPR 2026 · Table 3
Visual and representation RL
233.22PlanLLM
Vicuna-7B
PlanLLM: Video Procedure Planning with Refinable Large Language Models
AAAI 2025 · Table 2
Agents, code and math
332.27SCHEMA w/ VEDiT
SCHEMA
VEDIT: Latent Prediction Architecture For Procedural Video Representation Learning
ICLR 2025 · Procedure planning table
Video segmentation and tracking
432.09SCHEMASCHEMA: State CHangEs MAtter for Procedure Planning in Instructional Videos
ICLR 2024 · Table 3
Agents, code and math
530.44MTIDMasked Temporal Interpolation Diffusion for Procedure Planning in Instructional Videos
ICLR 2025 · Table 3
Action recognition
627.34PKDD (stdPDPP)
PDPP
Procedure Knowledge Decoupled Distillation Strategy for Procedure Planning in Instructional Videos
AAAI 2025 · Table 2
Knowledge distillation
724.47RAP (w/o unannotated)RAP: Retrieval-Augmented Planner for Adaptive Procedure Planning in Instructional Videos
ECCV 2024 · Table 3
Action recognition
823.65Skip-PlanSkip-Plan: Procedure Planning in Instructional Videos via Condensed Action Space Learning
ICCV 2023 · Table 1
Action recognition
921.33PDPPPDPP:Projected Diffusion for Procedure Planning in Instructional Videos
CVPR 2023 · Table 4
Action recognition
1020.25KEPP P2KG (R=2)Why Not Use Your Textbook? Knowledge-Enhanced Procedure Planning of Instructional Videos
CVPR 2024 · Table 4
Human-AI and goal-conditioned
2 more papers on this leaderboard, the baselines reported on it and a CSV download are free with a scime account.
Sign in free to see all

Not affiliated with any conference, journal or the benchmark's authors. Corrections: contact@latenticlabs.com.