# Inference Time on Saliency Prediction SOTA comparison (test)

Source: [Inference Time on Saliency Prediction SOTA comparison (test)](<https://www.sota2.com/research/sota/inference-time-on-saliency-prediction-sota-comparison-test>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2026-02-26T05:47:09.506Z

Evaluation of model inference time (latency) for Saliency Prediction models tested on the Saliency Prediction SOTA comparison (test) split.

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

| Scope | Recorded value |
| --- | --- |
| Task | Saliency Prediction |
| Dataset (including any stated split) | Saliency Prediction SOTA comparison (test) |
| Metric | Inference Time (s) |
| Unit | TIME |
| Direction | Lower is better |
| Results loaded | 7 of 7 |
| Coverage | All results currently returned for this benchmark |

Evaluation settings may differ between results. The original papers and recorded configurations are needed to establish comparability. A shared metric name alone does not establish an identical protocol.

## Metrics

- [Inference Time (s)](<https://www.sota2.com/research/sota/inference-time-on-saliency-prediction-sota-comparison-test.md?metric=Inference%20Time%20(s)>): Lower is better; unit: TIME.

## Recorded results

| Method | Inference Time (s) | Result date | Report paper |
| --- | --- | --- | --- |
| DINet | 0.02 | Not recorded | 68fa65c11e209131cea0c3cc |
| DVA | 0.02 | Not recorded | 68fa65c11e209131cea0c3cc |
| DINet | 0.03 | Not recorded | 68fa65c11e209131cea0c3cc |
| DINet | 0.06 | Not recorded | 68fa65c11e209131cea0c3cc |
| SAM | 0.07 | Not recorded | 68fa65c11e209131cea0c3cc |
| SAM | 0.09 | Not recorded | 68fa65c11e209131cea0c3cc |
| DSCLRCN | 0.27 | Not recorded | 68fa65c11e209131cea0c3cc |

### DINet

Inference Time (s): 0.02. Result ID: 698cc7748b517eb4331caead.

[Reporting paper](<https://www.sota2.com/research/paper/68fa65c11e209131cea0c3cc.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa65c11e209131cea0c3cc.md>)

[Implementation](<https://github.com/ysyscool/DINet>)

Recorded evaluation configuration: {"Backbone network":"Dilated ResNet50","#parameters":"27.04M","key module":"Dilated inception module for multi-scale","Input image size":"240 x 320"}. This record does not assert independent reproduction.

### DVA

Inference Time (s): 0.02. Result ID: 698cc7748b517eb4331caeb3.

[Reporting paper](<https://www.sota2.com/research/paper/68fa65c11e209131cea0c3cc.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa64271e209131cea02ef4.md>)

Recorded evaluation configuration: {"Backbone network":"VGG16","#parameters":"25.07M","key module":"Skip-layers for multi-scale","Input image size":"224 x 224"}. This record does not assert independent reproduction.

### DINet

Inference Time (s): 0.03. Result ID: 698cc7748b517eb4331caeae.

[Reporting paper](<https://www.sota2.com/research/paper/68fa65c11e209131cea0c3cc.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa65c11e209131cea0c3cc.md>)

[Implementation](<https://github.com/ysyscool/DINet>)

Recorded evaluation configuration: {"Backbone network":"Dilated ResNet50","#parameters":"27.04M","key module":"Dilated inception module for multi-scale","Input image size":"320 x 480"}. This record does not assert independent reproduction.

### DINet

Inference Time (s): 0.06. Result ID: 698cc7748b517eb4331caeaf.

[Reporting paper](<https://www.sota2.com/research/paper/68fa65c11e209131cea0c3cc.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa65c11e209131cea0c3cc.md>)

[Implementation](<https://github.com/ysyscool/DINet>)

Recorded evaluation configuration: {"Backbone network":"Dilated ResNet50","#parameters":"27.04M","key module":"Dilated inception module for multi-scale","Input image size":"480 x 640"}. This record does not assert independent reproduction.

### SAM

Inference Time (s): 0.07. Result ID: 698cc7748b517eb4331caeb1.

[Reporting paper](<https://www.sota2.com/research/paper/68fa65c11e209131cea0c3cc.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa64851e209131cea051f1.md>)

Recorded evaluation configuration: {"Backbone network":"Dilated VGG16","#parameters":"51.84M","key module":"Conv-LSTMs \[56\] for iterative refinement","Input image size":"240 x 320"}. This record does not assert independent reproduction.

### SAM

Inference Time (s): 0.09. Result ID: 698cc7748b517eb4331caeb0.

[Reporting paper](<https://www.sota2.com/research/paper/68fa65c11e209131cea0c3cc.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa64851e209131cea051f1.md>)

Recorded evaluation configuration: {"Backbone network":"Dilated ResNet50","#parameters":"70.09M","key module":"Conv-LSTMs \[56\] for iterative refinement","Input image size":"240 x 320"}. This record does not assert independent reproduction.

### DSCLRCN

Inference Time (s): 0.27. Result ID: 698cc7748b517eb4331caeb2.

[Reporting paper](<https://www.sota2.com/research/paper/68fa65c11e209131cea0c3cc.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa62d41e209131ce9fb68f.md>)

Recorded evaluation configuration: {"Backbone network":"Dilated ResNet50 + Places-CNN","#parameters":"\>33.71M","key module":"Spatial LSTMs \[57\] for context incorporation","Input image size":"480 x 640 + 227 x 227"}. This record does not assert independent reproduction.

Page 1 of 1. This is an index page, not the full collection.