Pith. sign in

REVIEW 3 major objections 5 minor 78 references

SPGNet: Semantic Prediction Guidance for Scene Parsing

T0 review · 3 major / 5 minor · reviewed 2026-08-14 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read Re-weighting local features by a mask derived from pixel-wise semantic predictions lets a two-stage encoder-decoder network beat a one-stage network at similar parameters and computation; on Cityscapes the resulting SPGNet reaches 81.1…

desk verdict SPG is a simple, effective supervise-and-excite attention module for two-stage segmentation, but the paper's matched-budget efficiency claim is not actually supported by its own tables. read the letter →

arxiv 1908.09798 v1 pith:HU5R3DSF submitted 2019-08-26 cs.CV

Bowen Cheng , Liang-Chieh Chen , Yunchao Wei , Yukun Zhu , Zilong Huang , Jinjun Xiong , Thomas Huang , Wen-Mei Hwu
show 1 more author
Honghui Shi
This is my paper · ORCID
classification cs.CV
keywords semanticsegmentationsceneparsingmulti-stageencoder-decoderattentionmodulefeaturere-weightingpixel-wisesupervisionCityscapessupervise-and-excite
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper tries to show that multi-stage encoder-decoder networks, long used in human pose estimation, can be made effective for semantic segmentation if the features passed between stages are re-weighted by a learned mask derived from pixel-wise semantic predictions. The proposed Semantic Prediction Guidance (SPG) module injects supervision into a first-stage decoder output, converts that prediction into a per-pixel guided attention mask, and uses the mask to excite local features before they enter the next stage. The authors argue that a two-stage SPGNet significantly outperforms its one-stage counterpart with similar parameters and computations, reporting 81.1 mean IoU on the Cityscapes test set using only fine annotations. If true, this offers a cheap way to capture contextual information without the heavy atrous convolutions used by many state-of-the-art segmentation models.

What carries the argument

The load-bearing mechanism is the Semantic Prediction Guidance (SPG) module, a supervise-and-excite block. It differs from squeeze-and-excite and gather-and-excite by supervising an intermediate prediction and using that prediction, after a 1x1 convolution and sigmoid, to generate a per-pixel, per-channel Guided Attention mask; the mask re-weights a transformed decoder feature map before a residual-style identity addition feeds the next encoder. The module is cheap, adding about 1% extra FLOPs and 2.3% higher peak memory, and its attention maps can be visualized by projecting the class-to-channel weights, which the authors link to object localization, co-occurrence, and the separation of semantically similar classes. The surrounding machinery is a stacked hourglass-style architecture with feature-pyramid encoders, a global-average-pooling image-level feature, an upsample module that fuses encoder and decoder features, and cross-stage feature aggregation borrowed from pose-estimation work.

What would settle it

Train the two-stage baseline with the same auxiliary pixel-wise semantic loss on the first-stage decoder output but with no guided-attention mask, meaning no 1x1-conv-plus-sigmoid re-weighting; if its validation mIoU reaches 77.67 or higher, the SPG mask contributes little beyond supervision. Table 3 of the paper does not include this condition.

Watch

Extended reading notes

Core claim

The central discovery is that semantic predictions themselves can serve as an attention signal: in SPGNet, the first-stage decoder output is passed through a 1x1 convolution to produce per-class logits, which are supervised during training, and a second 1x1 convolution with sigmoid turns those logits into a Guided Attention mask in feature-channel space. This mask is element-wise multiplied with a transformed version of the decoder features, and an identity path adds the original features back, producing the input to the next stage's encoder. On Cityscapes validation, the full SPG (sigmoid with identity and supervision) reaches 77.67 mIoU with a two-stage ResNet-18 configuration, versus 76.31 for the two-stage baseline without SPG; with ResNet-50 encoders, decoder width 256, and online hard example mining, the model reaches 80.91 mIoU on validation and 81.1 on the test set with multi-scale inference. The paper interprets this as evidence that a carefully re-weighted two-stage design can capture context more cheaply than single-stage networks with large atrous receptive fields.

Load-bearing premise

The paper attributes the gain to the SPG re-weighting mask, but its ablation never compares the two-stage baseline with the added pixel-wise supervision alone, so the supervision, not the attention mechanism, could be responsible for much of the improvement.

Editorial extensions

If this is right

  • A two-stage encoder-decoder with SPG outperforms a one-stage network with similar parameters and computations, so multi-stage designs can be competitive in segmentation without adding heavy context modules.
  • On Cityscapes, SPGNet reaches 81.1 mIoU on the test set using only fine annotations, beating DenseASPP by 0.5 mIoU while using about half its computation, and matching DANet closely at 22.7% of its FLOPs.
  • The supervision in SPG makes the re-weighting interpretable: guided attention maps localize objects, respond to co-occurring context, and can distinguish similar classes, which is useful for debugging and analysis.
  • The design transfers: a two-encoder SPGNet reaches 77.33 mIoU on PASCAL VOC 2012 and 67.23 mIoU (68.36 with multi-scale) on PASCAL Person-Part without extra pose data.
  • Adding more than two stages does not help on Cityscapes, since a three-stage version reaches 77.66 mIoU versus 77.67 for two stages, so the benefit saturates immediately.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • The ablation table never tests pixel-wise auxiliary supervision alone on the two-stage baseline; a reasonable inference, testable by adding only the stage-1 loss to the baseline without the guided-attention mask, is that much of the 1.36 mIoU gain could come from the extra supervision rather than from the re-weighting.
  • Since the gather step is implicitly done by the encoder-decoder, SPG could be inserted between any two feature hierarchies with rich multi-scale information, not only stacked encoder-decoders; testing it between DeepLab-style or FPN-style levels would show whether the mechanism is tied to the hourglass shape.
  • The guided-attention visualization suggests the mask is learning something close to an objectness or co-occurrence prior; if that holds, the same module could be applied to other dense prediction tasks such as depth estimation or panoptic segmentation, where a cheap attention signal from a coarse prediction might refine boundaries.
  • Given that SPGNet's failure classes (truck, bus, train) activate together in the attention maps, a testable extension would be to suppress guidance from aggregating over mutually confusable classes, for example by adding a confusion-aware term to the supervision.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 5 minor

Summary. The paper proposes Semantic Prediction Guidance (SPG), an attention module for semantic segmentation that re-weights decoder features using a supervised per-pixel prediction from an earlier stage, and embeds it in a two-stage encoder-decoder network (SPGNet). On Cityscapes test with only fine annotations, SPGNet reaches 81.1 mIoU with 654.8B FLOPs and 59.8M parameters, comparing favorably against DenseASPP and DANet in terms of efficiency. The paper reports extensive ablations on module design, number of stages, backbone combination, encoder depth, OHEM, decoder channels, and visualizations of the learned attention, plus generalization results on PASCAL VOC 2012 and PASCAL Person-Part.

Significance. The SPG module is simple, and the 'supervise-and-excite' idea is a reasonable extension of SE/GE with an intuitive interpretation. The validation ablations are extensive and the attention visualizations are informative. If the central claim about matching a one-stage counterpart at similar parameters and computation were cleanly supported, the efficiency argument would be a useful contribution. However, the headline claim is not established by the current tables, and the attribution of the gains to the guidance mechanism is confounded by the auxiliary supervision. The paper is likely worth publishing after additional controlled experiments.

major comments (3)
  1. [§4.4, Table 7] The central claim that a two-stage SPGNet 'significantly outperforms its one-stage counterpart with similar parameters and computations' is not supported by the closest matched-budget comparison in Table 7: one-stage ResNet-50 (24.7M params, 212.9B FLOPs, 77.80 mIoU) outperforms two-stage ResNet-18 with SPG (23.9M, 218.0B, 77.67 mIoU) by 0.13 mIoU. The alternative comparison favoring the two-stage model, two-stage ResNet-50 (55.6M, 467.6B, 79.81) versus one-stage ResNet-152 (59.4M, 530.1B, 78.33), uses decoder width 256 and OHEM for the former but width 128 and no OHEM for the latter, and the one-stage ResNet-152 result is lower than the one-stage ResNet-101 result (78.72) in the same table, which suggests an uncontrolled variable or instability. A clean experiment that fixes total budget and training protocol while toggling stage count and the SPG module is needed to substantiate the abstract and Section 5 conclusions.
  2. [§4.4, Table 3] The ablation does not isolate the effect of the semantic supervision from the attention mechanism. The baseline two-stage encoder-decoder with CSFA and no SPG is 76.31 mIoU; SPG without supervision is 77.12; SPG with supervision is 77.67. There is no run that adds only the per-stage auxiliary loss to the baseline without the attention mask, so the reported 1.36-point gain relative to baseline could be largely due to the extra supervision rather than to the 'excite' re-weighting. The authors' own statement that 'the additional gain mainly comes from adding supervision' (in the SE/GE comparison of Table 4) makes this confound central rather than peripheral.
  3. [§4.4, Tables 7 and 9] The final model configuration (2-stage ResNet-50, decoder channels=256, OHEM, multi-scale inference) is not the configuration used in the matched-budget ablations of Table 7, which use decoder channels=128 and no OHEM. Since Table 9 shows that increasing decoder channels from 128 to 256 adds 1.10 mIoU for ResNet-50 with OHEM, the margins in Table 7 cannot be directly used to predict the behavior of the final model. The authors should report one-stage counterparts at the same decoder width and with the same OHEM setting when making the claimed efficiency comparison.
minor comments (5)
  1. [Table 4 caption] The caption says 'SGP' instead of 'SPG'.
  2. [§4.3 and Table 2 caption] The text states that SPGNet outperforms DenseASPP in 15 out of 19 classes and DANet in 14 out of 19 classes, but the Table 2 caption says 'SPGNet outperforms existing top approaches in 13 out of 19 classes'; these numbers should be reconciled.
  3. [Table 10] The symbols in Table 10 for 'Extra data' and 'Multi-scale' appear inconsistent with the text: the text states 67.23 mIoU for single-scale and 68.36 for multi-scale, but the table's symbol pattern suggests the opposite. Please verify the table encoding.
  4. [§4.4, Table 5] The difference between 2-stage (77.67) and 3-stage (77.66) is within typical run-to-run variation; the conclusion that performance 'saturates' at 2 stages should be stated with awareness of this noise, ideally with multiple seeds.
  5. [§4.2] The paper does not state the random seed or number of repetitions for the ablation experiments; given the small margins in some comparisons, this information is important for reproducibility.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the SPG mechanism is a trained attention module and the reported results are benchmarked on held-out Cityscapes data, not derived from the method's inputs by construction.

full rationale

The paper's central claim is that a two-stage encoder-decoder with the SPG module outperforms its one-stage counterpart at similar parameter/computation budgets, and that SPGNet achieves 81.1 mIoU on Cityscapes test. This is an empirical claim. The Guided Attention in the SPG module is produced by a 1x1 convolution with sigmoid applied to the first stage's pixel-wise prediction, and the module is trained with real semantic labels; the supervision is a training signal rather than a constant fitted to reproduce the test numbers. The final test results are evaluated on a held-out dataset and compared against external baselines (DenseASPP, DANet), so the headline numbers are not derivable from the paper's definitions. The ablation tables (Table 3, Table 4) compare variants with and without the SPG module, supervision, identity mapping, and with SE/GE modules, which provides direct evidence about the module's contribution rather than a circular reduction. The only load-bearing external citation is Cross Stage Feature Aggregation [33], which is an architectural component imported from prior work and used as an off-the-shelf design choice; it is not a self-citation and its inclusion does not define the reported gains into existence. No equation in the paper equates a predicted quantity to a fitted input, and no 'uniqueness theorem' or author-derived prior result is invoked to forbid alternatives. The concern that the improvement over the one-stage baseline may be partly attributable to added supervision or increased capacity rather than the attention mask is a valid experimental-design caveat, but it is not circularity: it questions whether the measured gain is correctly attributed, not whether the result is equivalent to its inputs by construction. Accordingly, the circularity score is 0.

Assumptions & free parameters 6 free parameters · 4 assumptions · 0 invented entities

The central claim rests on standard deep learning assumptions and on several architecture choices tuned on the Cityscapes validation set. The main unstated cost is that these choices, including stage count, decoder width, activation, OHEM, and inference scales, are selected after observing validation performance, which is a mild form of fitting to the benchmark.

free parameters (6)
  • Number of encoder-decoder stages = 2
    Selected by validation ablation (Table 5): 1 stage 74.48, 2 stages 77.67, 3 stages 77.66. The headline model uses 2 stages.
  • Decoder channel width = 256
    Selected by validation ablation (Table 9): ResNet-50 with 256 channels gives 80.91 versus 80.10 with 128 channels, and this setting is kept for the final ResNet-50 model.
  • SPG excitation activation = sigmoid
    Chosen over softmax and sum variants by validation mIoU (Table 3): sigmoid 77.67, softmax 77.17, sum 76.96.
  • OHEM usage = enabled
    Ablation in Table 8 shows OHEM improves val mIoU (79.81 to 80.10 for ResNet-50, 80.04 to 80.85 for ResNet-101) and is used in the final model.
  • Multi-scale inference scales = 0.75, 1.0, 1.25, 1.5, 1.75, 2.0
    Added after the best val model, improving mIoU from 80.91 to 81.86; the test-set 81.1 is reported with this strategy.
  • Backbone combination = ResNet-50 + ResNet-50
    Ablation in Table 6 selects the two-stage encoder combination with best val mIoU (79.81) among ResNet-18 and ResNet-50 mixes.
assumptions (4)
  • domain assumption Pretrained ImageNet weights provide a useful initialization for the encoder modules.
    All encoders are ImageNet-pretrained ResNets (Section 4.2); without this, the architecture could not be trained as described from scratch.
  • domain assumption Validation mIoU on Cityscapes is a reliable proxy for test performance and for generalization to other datasets.
    Final model and hyperparameters are selected on the validation split (Section 4.4) and then reported on test (Table 1) and on PASCAL datasets (Section 4.6).
  • domain assumption Published benchmark numbers from other models are comparable despite differing training protocols and inference details.
    Table 1 compares mIoU across papers with different backbones, training splits, and multi-scale inference, assuming the reported numbers are directly comparable.
  • standard math Backpropagation and standard stochastic optimization are valid for training the proposed module.
    The paper relies on standard supervised training of differentiable CNNs; no new optimization theory is introduced.

how reviews work

0 comments
Cite this review

Pith. "Pith review of SPGNet: Semantic Prediction Guidance for Scene Parsing." pith.science (2026). https://pith.science/paper/HU5R3DSF

@misc{pith2026190809798,
  author       = {Pith},
  title        = {Pith review of: SPGNet: Semantic Prediction Guidance for Scene Parsing},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/HU5R3DSF}},
  note         = {Machine review of arXiv:1908.09798}
}
read the original abstract

Multi-scale context module and single-stage encoder-decoder structure are commonly employed for semantic segmentation. The multi-scale context module refers to the operations to aggregate feature responses from a large spatial extent, while the single-stage encoder-decoder structure encodes the high-level semantic information in the encoder path and recovers the boundary information in the decoder path. In contrast, multi-stage encoder-decoder networks have been widely used in human pose estimation and show superior performance than their single-stage counterpart. However, few efforts have been attempted to bring this effective design to semantic segmentation. In this work, we propose a Semantic Prediction Guidance (SPG) module which learns to re-weight the local features through the guidance from pixel-wise semantic prediction. We find that by carefully re-weighting features across stages, a two-stage encoder-decoder network coupled with our proposed SPG module can significantly outperform its one-stage counterpart with similar parameters and computations. Finally, we report experimental results on the semantic segmentation benchmark Cityscapes, in which our SPGNet attains 81.1% on the test set using only 'fine' annotations.

Figures

Figures reproduced from arXiv: 1908.09798 by the authors.

Figure 1
Figure 1. Three different frameworks for re-weighting local fea [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. The overall structure of SPGNet. Only two stages are shown for simplicity and it can be easily generalized to more stages. (a) [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. Our Semantic Prediction Guidance (SPG) module. [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figures from the paper (4 more)
Figure 4
Figure 4. Figure 4: Method to visualize guided attention. OHEM consistently improves the performance. Decoder channels. We experiment on the effect of decoder channels in [PITH_FULL_IMAGE:figures/full_fig_p007_4.png]
Figure 5
Figure 5. Figure 5: Visualization of guided attention for 4 general classes. Guided attention focuses on the boundary of co-occurred objects/things. [PITH_FULL_IMAGE:figures/full_fig_p008_5.png]
Figure 6
Figure 6. Figure 6: Visualization of guided attention for Person/Rider. Guided attention is capable of differentiating semantically similar classes. [PITH_FULL_IMAGE:figures/full_fig_p008_6.png]
Figure 7
Figure 7. Figure 7: Visualization of guided attention for Truck/Bus/Train. Our failure cases where guided attention confuses among Truck/Bus/Train. [PITH_FULL_IMAGE:figures/full_fig_p008_7.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

78 extracted references · 64 canonical work pages

  1. [1]

    Gated feedback refinement network for dense image labeling

    Md Amirul Islam, Mrigank Rochan, Neil DB Bruce, and Yang Wang. Gated feedback refinement network for dense image labeling. In CVPR, 2017. 2

    work page 2017

  2. [2]

    2d human pose estimation: New benchmark and state of the art analysis

    Mykhaylo Andriluka, Leonid Pishchulin, Peter Gehler, and Bernt Schiele. 2d human pose estimation: New benchmark and state of the art analysis. In CVPR, 2014. 8

    work page 2014

  3. [3]

    Segnet: A deep convolutional encoder-decoder architecture for image segmentation

    Vijay Badrinarayanan, Alex Kendall, and Roberto Cipolla. Segnet: A deep convolutional encoder-decoder architecture for image segmentation. IEEE TPAMI, 2017. 2

    work page 2017

  4. [4]

    Loss maxpooling for semantic image segmentation

    Samuel Rota Bul `o, Gerhard Neuhold, and Peter Kontschieder. Loss maxpooling for semantic image segmentation. In CVPR, 2017. 6

    work page 2017

  5. [5]

    Realtime multi-person 2d pose estimation using part affinity fields

    Zhe Cao, Tomas Simon, Shih-En Wei, and Yaser Sheikh. Realtime multi-person 2d pose estimation using part affinity fields. In CVPR, 2017. 2

    work page 2017

  6. [6]

    Collins, Yukun Zhu, George Papandreou, Barret Zoph, Florian Schroff, Hartwig Adam, and Jonathon Shlens

    Liang-Chieh Chen, Maxwell D. Collins, Yukun Zhu, George Papandreou, Barret Zoph, Florian Schroff, Hartwig Adam, and Jonathon Shlens. Searching for efficient multi-scale ar- chitectures for dense image prediction. In NIPS, 2018. 2, 7

    work page 2018

  7. [7]

    Semantic image segmen- tation with deep convolutional nets and fully connected crfs

    Liang-Chieh Chen, George Papandreou, Iasonas Kokkinos, Kevin Murphy, and Alan L Yuille. Semantic image segmen- tation with deep convolutional nets and fully connected crfs. In ICLR, 2015. 1, 2

    work page 2015

  8. [8]

    Deeplab: Semantic image segmentation with deep convolutional nets, atrous convolu- tion, and fully connected crfs

    Liang-Chieh Chen, George Papandreou, Iasonas Kokkinos, Kevin Murphy, and Alan L Yuille. Deeplab: Semantic image segmentation with deep convolutional nets, atrous convolu- tion, and fully connected crfs. IEEE TPAMI, 2017. 2

    work page 2017

Show all 78 references
  1. [9]

    Rethinking atrous convolution for seman- tic image segmentation

    Liang-Chieh Chen, George Papandreou, Florian Schroff, and Hartwig Adam. Rethinking atrous convolution for seman- tic image segmentation. arXiv preprint arXiv:1706.05587 ,

    arXiv

  2. [10]

    Attention to scale: Scale-aware semantic im- age segmentation

    Liang-Chieh Chen, Yi Yang, Jiang Wang, Wei Xu, and Alan L Yuille. Attention to scale: Scale-aware semantic im- age segmentation. In CVPR, 2016. 3 9

    2016

  3. [11]

    Encoder-decoder with atrous separable convolution for semantic image segmentation

    Liang-Chieh Chen, Yukun Zhu, George Papandreou, Florian Schroff, and Hartwig Adam. Encoder-decoder with atrous separable convolution for semantic image segmentation. In ECCV, 2018. 2, 8

    2018

  4. [12]

    Detect what you can: De- tecting and representing objects using holistic models and body parts

    Xianjie Chen, Roozbeh Mottaghi, Xiaobai Liu, Sanja Fidler, Raquel Urtasun, and Alan Yuille. Detect what you can: De- tecting and representing objects using holistic models and body parts. In CVPR, pages 1971–1978, 2014. 7

    1971

  5. [13]

    Aˆ 2-nets: Double attention networks

    Yunpeng Chen, Yannis Kalantidis, Jianshu Li, Shuicheng Yan, and Jiashi Feng. Aˆ 2-nets: Double attention networks. In NIPS, 2018. 3

    2018

  6. [14]

    The cityscapes dataset for semantic urban scene understanding

    Marius Cordts, Mohamed Omran, Sebastian Ramos, Timo Rehfeld, Markus Enzweiler, Rodrigo Benenson, Uwe Franke, Stefan Roth, and Bernt Schiele. The cityscapes dataset for semantic urban scene understanding. In CVPR,

  7. [15]

    Imagenet: A large-scale hierarchical image database

    Jia Deng, Wei Dong, Richard Socher, Li-Jia Li, Kai Li, and Li Fei-Fei. Imagenet: A large-scale hierarchical image database. In CVPR, 2009. 4

    2009

  8. [16]

    Everingham, L

    M. Everingham, L. Van Gool, C. K. I. Williams, J. Winn, and A. Zisserman. The pascal visual object classes (voc) challenge. IJCV, 2010. 1, 7

    2010

  9. [17]

    Weakly and semi supervised human body part parsing via pose-guided knowledge trans- fer

    Hao-Shu Fang, Guansong Lu, Xiaolin Fang, Jianwen Xie, Yu-Wing Tai, and Cewu Lu. Weakly and semi supervised human body part parsing via pose-guided knowledge trans- fer. arXiv preprint arXiv:1805.04310, 2018. 7, 8

    2018 arXiv

  10. [18]

    Dual attention network for scene segmentation

    Jun Fu, Jing Liu, Haijie Tian, Yong Li, Yongjun Bao, Zhiwei Fang, and Hanqing Lu. Dual attention network for scene segmentation. In CVPR, 2019. 2, 3, 5, 6

    2019

  11. [19]

    Stacked deconvolutional network for se- mantic segmentation

    Jun Fu, Jing Liu, Yuhang Wang, Jin Zhou, Changyong Wang, and Hanqing Lu. Stacked deconvolutional network for se- mantic segmentation. IEEE TIP, 2019. 2

    2019

  12. [20]

    Deep residual learning for image recognition

    Kaiming He, Xiangyu Zhang, Shaoqing Ren, and Jian Sun. Deep residual learning for image recognition. In CVPR,

  13. [21]

    Zemel, and Miguel ´A

    Xuming He, Richard S. Zemel, and Miguel ´A. Carreira- Perpi˜n´an. Multiscale conditional random fields for image labeling. In CVPR, 2004. 1

    2004

  14. [22]

    A real-time algorithm for signal analysis with the help of the wavelet transform

    Matthias Holschneider, Richard Kronland-Martinet, Jean Morlet, and Ph Tchamitchian. A real-time algorithm for signal analysis with the help of the wavelet transform. In Wavelets: Time-Frequency Methods and Phase Space, pages 289–297. 1989. 1, 2

    1989

  15. [23]

    Relation networks for object detection

    Han Hu, Jiayuan Gu, Zheng Zhang, Jifeng Dai, and Yichen Wei. Relation networks for object detection. InCVPR, 2018. 3

    2018

  16. [24]

    Gather-excite: Exploiting feature context in convo- lutional neural networks

    Jie Hu, Li Shen, Samuel Albanie, Gang Sun, and Andrea Vedaldi. Gather-excite: Exploiting feature context in convo- lutional neural networks. In NIPS, 2018. 1, 2, 3, 4, 5, 6

    2018

  17. [25]

    Squeeze-and-excitation net- works

    Jie Hu, Li Shen, and Gang Sun. Squeeze-and-excitation net- works. In CVPR, 2018. 1, 2, 3, 4, 5, 6

    2018

  18. [26]

    Ccnet: Criss-cross attention for semantic segmentation

    Zilong Huang, Xinggang Wang, Lichao Huang, Chang Huang, Yunchao Wei, and Wenyu Liu. Ccnet: Criss-cross attention for semantic segmentation. In ICCV, 2019. 3

    2019

  19. [27]

    Geometry-aware dis- tillation for indoor semantic segmentation

    Jianbo Jiao, Yunchao Wei, Zequn Jie, Honghui Shi, Ryn- son WH Lau, and Thomas S Huang. Geometry-aware dis- tillation for indoor semantic segmentation. In CVPR, 2019. 2

    2019

  20. [28]

    Multi-scale structure-aware network for human pose estimation

    Lipeng Ke, Ming-Ching Chang, Honggang Qi, and Siwei Lyu. Multi-scale structure-aware network for human pose estimation. In ECCV, 2018. 2

    2018

  21. [29]

    Recurrent scene parsing with perspective understanding in the loop

    Shu Kong and Charless C Fowlkes. Recurrent scene parsing with perspective understanding in the loop. In CVPR, 2018. 3, 5

    2018

  22. [30]

    Imagenet classification with deep convolutional neural net- works

    Alex Krizhevsky, Ilya Sutskever, and Geoffrey E Hinton. Imagenet classification with deep convolutional neural net- works. In NIPS, 2012. 1, 2

    2012

  23. [31]

    What, where and how many? com- bining object detectors and crfs

    L’ubor Ladick `y, Paul Sturgess, Karteek Alahari, Chris Rus- sell, and Philip HS Torr. What, where and how many? com- bining object detectors and crfs. In ECCV, 2010. 1

    2010

  24. [32]

    Backpropagation applied to handwrit- ten zip code recognition.Neural computation, 1(4):541–551,

    Yann LeCun, Bernhard Boser, John S Denker, Donnie Henderson, Richard E Howard, Wayne Hubbard, and Lawrence D Jackel. Backpropagation applied to handwrit- ten zip code recognition.Neural computation, 1(4):541–551,

  25. [33]

    Rethinking on multi-stage networks for human pose estimation

    Wenbo Li, Zhicheng Wang, Binyi Yin, Qixiang Peng, Yum- ing Du, Tianzi Xiao, Gang Yu, Hongtao Lu, Yichen Wei, and Jian Sun. Rethinking on multi-stage networks for human pose estimation. arXiv preprint arXiv:1901.00148, 2019. 2, 3, 5

    1901 arXiv

  26. [34]

    Not all pixels are equal: Difficulty-aware se- mantic segmentation via deep layer cascade

    Xiaoxiao Li, Ziwei Liu, Ping Luo, Chen Change Loy, and Xiaoou Tang. Not all pixels are equal: Difficulty-aware se- mantic segmentation via deep layer cascade. InCVPR, 2017. 2

    2017

  27. [35]

    Interpretable structure- evolving lstm

    Xiaodan Liang, Liang Lin, Xiaohui Shen, Jiashi Feng, Shuicheng Yan, and Eric P Xing. Interpretable structure- evolving lstm. In CVPR, 2017. 7

    2017

  28. [36]

    Refinenet: Multi-path refinement networks for high- resolution semantic segmentation

    Guosheng Lin, Anton Milan, Chunhua Shen, and Ian Reid. Refinenet: Multi-path refinement networks for high- resolution semantic segmentation. In CVPR, 2017. 2, 5

    2017

  29. [37]

    Feature pyramid networks for object detection

    Tsung-Yi Lin, Piotr Doll ´ar, Ross Girshick, Kaiming He, Bharath Hariharan, and Serge Belongie. Feature pyramid networks for object detection. In CVPR, 2017. 4, 9

    2017

  30. [38]

    Auto-deeplab: Hierarchical neural architecture search for semantic image segmentation

    Chenxi Liu, Liang-Chieh Chen, Florian Schroff, Hartwig Adam, Wei Hua, Alan Yuille, and Li Fei-Fei. Auto-deeplab: Hierarchical neural architecture search for semantic image segmentation. arXiv preprint arXiv:1901.02985, 2019. 2

    1901 arXiv

  31. [39]

    Parsenet: Looking wider to see better

    Wei Liu, Andrew Rabinovich, and Alexander C Berg. Parsenet: Looking wider to see better. arXiv:1506.04579,

    arXiv

  32. [40]

    Fully convolutional networks for semantic segmentation

    Jonathan Long, Evan Shelhamer, and Trevor Darrell. Fully convolutional networks for semantic segmentation. In IEEE TPAMI, 2015. 1, 2

    2015

  33. [41]

    Associa- tive embedding: End-to-end learning for joint detection and grouping

    Alejandro Newell, Zhiao Huang, and Jia Deng. Associa- tive embedding: End-to-end learning for joint detection and grouping. In NIPS, 2017. 2

    2017

  34. [42]

    Stacked hour- glass networks for human pose estimation

    Alejandro Newell, Kaiyu Yang, and Jia Deng. Stacked hour- glass networks for human pose estimation. In ECCV, 2016. 2, 3, 4

    2016

  35. [43]

    Learning deconvolution network for semantic segmentation

    Hyeonwoo Noh, Seunghoon Hong, and Bohyung Han. Learning deconvolution network for semantic segmentation. In ICCV, 2015. 2

    2015

  36. [44]

    Modeling local and global deformations in deep learning: Epitomic convolution, multiple instance learning, and sliding window detection

    George Papandreou, Iasonas Kokkinos, and Pierre-Andre Savalle. Modeling local and global deformations in deep learning: Epitomic convolution, multiple instance learning, and sliding window detection. In CVPR, 2015. 1, 2 10

    2015

  37. [45]

    Large kernel matters–improve semantic segmenta- tion by global convolutional network

    Chao Peng, Xiangyu Zhang, Gang Yu, Guiming Luo, and Jian Sun. Large kernel matters–improve semantic segmenta- tion by global convolutional network. In CVPR, 2017. 2

    2017

  38. [46]

    Full-resolution residual networks for seman- tic segmentation in street scenes

    Tobias Pohlen, Alexander Hermans, Markus Mathias, and Bastian Leibe. Full-resolution residual networks for seman- tic segmentation in street scenes. In CVPR, 2017. 2

    2017

  39. [47]

    Seamless scene segmentation

    Lorenzo Porzi, Samuel Rota Bulo, Aleksander Colovic, and Peter Kontschieder. Seamless scene segmentation. In CVPR,

  40. [48]

    U-net: Convolutional networks for biomedical image segmentation

    Olaf Ronneberger, Philipp Fischer, and Thomas Brox. U-net: Convolutional networks for biomedical image segmentation. In MICCAI. Springer, 2015. 2

    2015

  41. [49]

    In-place activated batchnorm for memory-optimized training of dnns

    Samuel Rota Bul `o, Lorenzo Porzi, and Peter Kontschieder. In-place activated batchnorm for memory-optimized training of dnns. In CVPR, 2018. 4

    2018

  42. [50]

    Overfeat: Integrated recognition, localization and detection using convolutional networks

    Pierre Sermanet, David Eigen, Xiang Zhang, Micha ¨el Math- ieu, Rob Fergus, and Yann LeCun. Overfeat: Integrated recognition, localization and detection using convolutional networks. arXiv preprint arXiv:1312.6229, 2013. 1, 2

    2013 arXiv

  43. [51]

    Stacked u-nets: a no-frills approach to natural image segmentation

    Sohil Shah, Pallabi Ghosh, Larry S Davis, and Tom Gold- stein. Stacked u-nets: a no-frills approach to natural image segmentation. arXiv preprint arXiv:1804.10343, 2018. 2

    2018 arXiv

  44. [52]

    Very deep convo- lutional networks for large-scale image recognition

    Karen Simonyan and Andrew Zisserman. Very deep convo- lutional networks for large-scale image recognition. InICLR,

  45. [53]

    Spg-net: Segmentation prediction and guidance network for image inpainting

    Yuhang Song, Chao Yang, Yeji Shen, Peng Wang, Qin Huang, and C-C Jay Kuo. Spg-net: Segmentation prediction and guidance network for image inpainting. arXiv preprint arXiv:1805.03356, 2018. 2

    2018 arXiv

  46. [54]

    Going deeper with convolutions

    Christian Szegedy, Wei Liu, Yangqing Jia, Pierre Sermanet, Scott Reed, Dragomir Anguelov, Dumitru Erhan, Vincent Vanhoucke, and Andrew Rabinovich. Going deeper with convolutions. In CVPR, 2015. 2

    2015

  47. [55]

    Finding things: Image parsing with regions and per-exemplar detectors

    Joseph Tighe and Svetlana Lazebnik. Finding things: Image parsing with regions and per-exemplar detectors. In CVPR,

  48. [56]

    Image parsing: Unifying segmentation, detec- tion, and recognition

    Zhuowen Tu, Xiangrong Chen, Alan L Yuille, and Song- Chun Zhu. Image parsing: Unifying segmentation, detec- tion, and recognition. IJCV, 2005. 1

    2005

  49. [57]

    Attention is all you need

    Ashish Vaswani, Noam Shazeer, Niki Parmar, Jakob Uszko- reit, Llion Jones, Aidan N Gomez, Łukasz Kaiser, and Illia Polosukhin. Attention is all you need. In NIPS, 2017. 3

    2017

  50. [58]

    Understanding convolution for semantic segmentation

    Panqu Wang, Pengfei Chen, Ye Yuan, Ding Liu, Zehua Huang, Xiaodi Hou, and Garrison Cottrell. Understanding convolution for semantic segmentation. In WACV, 2018. 5

    2018

  51. [59]

    A stagewise refinement model for detecting salient objects in images

    Tiantian Wang, Ali Borji, Lihe Zhang, Pingping Zhang, and Huchuan Lu. A stagewise refinement model for detecting salient objects in images. In ICCV, 2017. 2

    2017

  52. [60]

    Non-local neural networks

    Xiaolong Wang, Ross Girshick, Abhinav Gupta, and Kaim- ing He. Non-local neural networks. In CVPR, 2018. 3

    2018

  53. [61]

    Convolutional pose machines

    Shih-En Wei, Varun Ramakrishna, Takeo Kanade, and Yaser Sheikh. Convolutional pose machines. In CVPR, 2016. 2

    2016

  54. [62]

    The devil is in the decoder: Classification, regression and gans

    Zbigniew Wojna, Vittorio Ferrari, Sergio Guadarrama, Nathan Silberman, Liang-Chieh Chen, Alireza Fathi, and Jasper Uijlings. The devil is in the decoder: Classification, regression and gans. IJCV, pages 1–13, 2019. 2

    2019

  55. [63]

    Bridging category-level and instance-level semantic image segmentation

    Zifeng Wu, Chunhua Shen, and Anton van den Hengel. Bridging category-level and instance-level semantic image segmentation. arXiv:1605.06885, 2016. 6

    2016 arXiv

  56. [64]

    Joint multi-person pose estimation and semantic part seg- mentation

    Fangting Xia, Peng Wang, Xianjie Chen, and Alan L Yuille. Joint multi-person pose estimation and semantic part seg- mentation. In CVPR, 2017. 7, 8

    2017

  57. [65]

    Unified perceptual parsing for scene understand- ing

    Tete Xiao, Yingcheng Liu, Bolei Zhou, Yuning Jiang, and Jian Sun. Unified perceptual parsing for scene understand- ing. In ECCV, 2018. 2

    2018

  58. [66]

    Denseaspp for semantic segmentation in street scenes

    Maoke Yang, Kun Yu, Chi Zhang, Zhiwei Li, and Kuiyuan Yang. Denseaspp for semantic segmentation in street scenes. In CVPR, 2018. 2, 5, 6

    2018

  59. [67]

    Deeperlab: Single-shot image parser

    Tien-Ju Yang, Maxwell D Collins, Yukun Zhu, Jyh-Jing Hwang, Ting Liu, Xiao Zhang, Vivienne Sze, George Pa- pandreou, and Liang-Chieh Chen. Deeperlab: Single-shot image parser. arXiv preprint arXiv:1902.05093, 2019. 2, 6

    1902 arXiv

  60. [68]

    Learning feature pyramids for human pose estimation

    Wei Yang, Shuang Li, Wanli Ouyang, Hongsheng Li, and Xiaogang Wang. Learning feature pyramids for human pose estimation. In ICCV, 2017. 2

    2017

  61. [69]

    Describing the scene as a whole: Joint object detection, scene classification and semantic segmentation

    Jian Yao, Sanja Fidler, and Raquel Urtasun. Describing the scene as a whole: Joint object detection, scene classification and semantic segmentation. In CVPR, 2012. 1

    2012

  62. [70]

    Bisenet: Bilateral segmenta- tion network for real-time semantic segmentation

    Changqian Yu, Jingbo Wang, Chao Peng, Changxin Gao, Gang Yu, and Nong Sang. Bisenet: Bilateral segmenta- tion network for real-time semantic segmentation. In ECCV,

  63. [71]

    Learning a discriminative feature network for semantic segmentation

    Changqian Yu, Jingbo Wang, Chao Peng, Changxin Gao, Gang Yu, and Nong Sang. Learning a discriminative feature network for semantic segmentation. In CVPR, 2018. 5

    2018

  64. [72]

    Con- text encoding for semantic segmentation

    Hang Zhang, Kristin Dana, Jianping Shi, Zhongyue Zhang, Xiaogang Wang, Ambrish Tyagi, and Amit Agrawal. Con- text encoding for semantic segmentation. In CVPR, 2018. 3

    2018

  65. [73]

    Scale-adaptive convolutions for scene pars- ing

    Rui Zhang, Sheng Tang, Yongdong Zhang, Jintao Li, and Shuicheng Yan. Scale-adaptive convolutions for scene pars- ing. In ICCV, 2017. 5

    2017

  66. [74]

    Exfuse: Enhancing feature fusion for semantic segmentation

    Zhenli Zhang, Xiangyu Zhang, Chao Peng, Xiangyang Xue, and Jian Sun. Exfuse: Enhancing feature fusion for semantic segmentation. In ECCV, 2018. 2

    2018

  67. [75]

    Pyramid scene parsing network

    Hengshuang Zhao, Jianping Shi, Xiaojuan Qi, Xiaogang Wang, and Jiaya Jia. Pyramid scene parsing network. In CVPR, 2017. 1, 2, 5, 9

    2017

  68. [76]

    Psanet: Point-wise spatial attention network for scene parsing

    Hengshuang Zhao, Yi Zhang, Shu Liu, Jianping Shi, Chen Change Loy, Dahua Lin, and Jiaya Jia. Psanet: Point-wise spatial attention network for scene parsing. In ECCV, 2018. 5

    2018

  69. [77]

    Scene parsing through ade20k dataset

    Bolei Zhou, Hang Zhao, Xavier Puig, Sanja Fidler, Adela Barriuso, and Antonio Torralba. Scene parsing through ade20k dataset. In CVPR, 2017. 1, 3

    2017

  70. [78]

    Barret Zoph and Quoc V . Le. Neural architecture search with reinforcement learning. In ICLR, 2017. 2 11

    2017

Pith tools

Reviewed August 14, 2026 · model on record in the stance chip above.