I looked at the paper and did not see where "performance" was defined or any hint as to what units it has. How is performance calculated? I see they compare "precision" to "recall".
From the article: “We evaluated the proposed method with other deep learning based methods and showed that our method results in faster and more accurate results,”
Therefore, I'd say that it's doing more than just rotating the image and re-running the algorithm, but you stick to doing that, it's cool.