Research map / Label-efficient and robust learning / Recognition and long-tail
Learning Deep Representation for Face Alignment with Auxiliary Attributes
Zhanpeng Zhang, Ping Luo, Chen Change Loy, Xiaoou Tang
This paper tried a common change: Add a reconstruction / auxiliary loss
add a random-objective auxiliary task · MAFL · mean error No clear effect
What 987 accepted papers report for this change
- 882 improved
- 41 no clear effect
- 20 mixed
- 44 hurt
- OPTION: Optimal Transport–Guided Flow Matching for Incomplete and Unaligned Multi-View Clustering ICML 2026 · add reconstruction loss L_rec · Scene-15 (aligned, 70% missing) · ACC: 32.4 -> 37.5Improved
- Detached Skip-Links and $R$-Probe: Decoupling Feature Aggregation from Gradient Propagation for MLLM OCR ICML 2026 · add R-probe auxiliary loss · OCRBench · score: 714.0 -> 721.0Improved
- Rethinking Time-Series Imputation as Conditional Inference along Temporal Evolution ICML 2026 · add reconstruction loss on observed entries (joint vs missing-only supervision) · ETTh1 10% missing · Masked MAE: 0.2002 -> 0.1658Improved
See this change in detailMatched by the wording of each paper's own result, read by AI from the full text. Papers report gains more often than failures, so this shows what gets published, not the chance a change works.
AI-extracted. The measured effects below were read by AI from the paper's full text (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 |
|---|---|---|---|
| add dynamic task coefficient instead of task-wise early stopping (no correlation learning) (ablation) | TCDCN · MAFL | mean error (x10^-2): 8.35 -> 8.07 | Improved |
| add dynamic task coefficient instead of task-wise early stopping (with correlation learning) (ablation) | TCDCN · MAFL | mean error (x10^-2): 8.21 -> 7.95 | Improved |
| add inter-task correlation learning with dynamic task coefficient (ablation) | TCDCN · MAFL | mean error (x10^-2): 8.07 -> 7.95 | Improved |
| add inter-task correlation learning with task-wise early stopping (ablation) | TCDCN · MAFL | mean error (x10^-2): 8.35 -> 8.21 | Improved |
| add dynamic task coefficient to training (ablation) | TCDCN · MAFL | training convergence | Improved |
| add all auxiliary attribute tasks (FLD+all) to landmark detection only (ablation) | TCDCN · MAFL | failure rate: over 7% | Improved |
| add head pose and global attribute groups as auxiliary tasks (ablation) | TCDCN · MAFL | mean error | Improved |
| add eyes or mouth attribute groups as auxiliary tasks (ablation) | TCDCN · MAFL | mean error | Improved |
| add a random-objective auxiliary task (ablation) | TCDCN · MAFL | mean error | No clear effect |
| use TCDCN instead of CFSS | TCDCN · Helen (194 landmarks) | mean error (%): 4.74 -> 4.63 | Improved |
| use TCDCN instead of CFSS | TCDCN · 300-W challenging subset | mean error (%): 9.98 -> 8.60 | Improved |
| use TCDCN instead of CFSS | TCDCN · 300-W common subset | mean error (%): 4.73 -> 4.80 | Hurt |
Closest papers by meaning
- Facial Landmark Detection with Tweaked Convolutional Neural Networks (TPAMI 2017 · similarity 0.90)
- Brute-Force Facial Landmark Analysis With a 140,000-Way Classifier (AAAI 2018 · similarity 0.88)
- Aggregation via Separation: Boosting Facial Landmark Detector With Semi-Supervised Style Translation (ICCV 2019 · similarity 0.87)
- How Far Are We From Solving the 2D & 3D Face Alignment Problem? (And a Dataset of 230,000 3D Facial Landmarks) (ICCV 2017 · similarity 0.87)
- Multi-Task Head Pose Estimation in-the-Wild (TPAMI 2020 · similarity 0.87)
- Attributes Shape the Embedding Space of Face Recognition Models (ICML 2025 · similarity 0.86)
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.
