Research map / Detection and segmentation / CNN features
Actor-Critic Instance Segmentation
Check your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| replace one-step BPTT baseline training with actor-critic training using future Dice rewards | recurrent instance segmentation · CVPPP val | SBD: 79.4 -> 80.5 | Helped |
| replace one-step BPTT baseline training with actor-critic training using future IoU rewards | recurrent instance segmentation · KITTI test | MUCov: 50.7 -> 57.3 | Helped |
| replace one-step BPTT baseline training with actor-critic training using future IoU rewards | recurrent instance segmentation · KITTI test | AvgFP: 0.393 -> 0.338 | Helped |
| replace one-step BPTT baseline training with actor-critic training using future IoU rewards | recurrent instance segmentation · KITTI test | AvgFN: 0.432 -> 0.309 | Helped |
| add the KL-divergence exploration term to the actor loss (ablation) | AC-Dice · CVPPP val | SBD: 75.4 -> 80.5 | Helped |
| add the KL-divergence exploration term to the actor loss (ablation) | AC-Dice · CVPPP val | |DiC|: 1.36 -> 0.88 | Helped |
| add the State Pyramid to the decoder (ablation) | AC-Dice · CVPPP val | SBD: 61.3 -> 80.5 | Helped |
| add the State Pyramid to the decoder (ablation) | AC-Dice · CVPPP val | |DiC|: 1.52 -> 0.88 | Helped |
| use truncated one-step BPTT instead of fully unrolled BPTT (ablation) | actor-only recurrent baseline · CVPPP val | SBD: 80.0 -> 79.4 | Hurt |
| remove the LSTM hidden state while retaining the accumulated mask (ablation) | AC-Dice · CVPPP val | Dice: 78.4 -> 78.5 | No clear effect |
| remove the accumulated mask while retaining the LSTM hidden state (ablation) | AC-Dice · CVPPP val | Dice: 78.4 -> 5.8 | Hurt |
| remove the LSTM hidden state while retaining the accumulated mask (ablation) | AC-Dice · CVPPP val | |DiC|: 0.88 -> 1.92 | Hurt |
Closest papers by meaning
- Reinforcement Cutting-Agent Learning for Video Object Segmentation (CVPR 2018 · similarity 0.88)
- Studying the Interplay Between the Actor and Critic Representations in Reinforcement Learning (ICLR 2025 · similarity 0.85)
- v-CLR: View-Consistent Learning for Open-World Instance Segmentation (CVPR 2025 · similarity 0.85)
- LENS: Learning to Segment Anything with Unified Reinforced Reasoning (AAAI 2026 · similarity 0.85)
- FalconPerception-HD: High Density Perception via Reinforcement Learning (NeurIPS 2026 · similarity 0.85)
- LookThere! Sparse Vision by Reinforced Selection (NeurIPS 2026 · similarity 0.85)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.