Semantic Segmentation via Structured Refined Prediction and Dual Global Priors
Bibliographic record
Abstract
Deep residual network and multi-scale context module become two key ingredients in recent researches which made great progress in semantic segmentation tasks. The multi-scale features utilized by these networks are also proven to be effective in refining object boundaries. However, regular convolution kernels are inherently limited to geometric transformation due to the fixed structure. Additionally, when integrating multi-scale information in boundary detection, the capability of global attention starts to degrade and the problem of inconsistent intra-class segmentation will occur. To solve above problems, we propose two novel architectures dubbed as Deformable Residual Boundary (DRB) module and Dual Global Prior (DGP) module. The DRB module can learn more refined boundary information from non-rigid geometric structures by deformable convolutions. The extracted information will be further embedded into the branches of an encoder-decoder architecture to handle the deformation of objects. For multi-scale fusion, the DGP module utilizes global priors to enhance the global attention, which alleviates the inconsistent intra-class segmentation problem especially when dealing with object surfaces of different textures/resolutions. We achieve a mean IOU value of 80.3% without coarse set, outperforming state-of-the-art approaches. When training with the coarse set, the final mIOU of the proposed method is 80.9%, also competitive to other approaches.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one teacher head, not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".