Created at 7am, Apr 19
SplinterSports
0
A Universal Protocol to Benchmark Camera Calibration for Sports
nfBGoX7wz9tQTbZuS65Y58ltfqQBuje4BiA4S0O8OaM
File Type
PDF
Entry Count
88
Embed. Model
jina_embeddings_v2_base_en
Index Type
hnsw

Camera calibration is a crucial component in the realm of sports analytics, as it serves as the foundation to extract 3D information out of the broadcast images. Despite the significance of camera calibration research in sports analytics, progress is impeded by outdated benchmarking criteria. Indeed, the annotation data and evaluation metrics provided by most currently available benchmarks strongly favor and incite the development of sports field registration methods, i.e. methods estimating homographies that map the sports field plane to the image plane. However, such homography-based methods are doomed to overlook the broader capabilities of camera calibration in bridging the 3D world to the image. In particular, real-world non-planar sports field elements (such as goals, corner flags, baskets, …) and image distortion caused by broadcast camera lenses are out of the scope of sports field registration methods. To overcome these limitations, we designed a new benchmarking protocol, named ProCC, based on two principles: (1) the protocol should be agnostic to the camera model chosen for a camera calibration method, and (2) the protocol should fairly evaluate camera calibration methods using the reprojection of arbitrary yet accurately known 3D objects. Indirectly, we also provide insights into the metric used in SoccerNet-calibration, which solely relies on image annotation data of viewed 3D objects as ground truth, thus implementing our protocol. With experiments on the World Cup 2014, CARWC, and SoccerNet datasets, we show that our benchmarking protocol provides fairer evaluations of camera calibration methods. By defining our requirements for proper benchmarking, we hope to pave the way for a new stage in camera calibration for sports applications with high accuracy standards.

Income Gini index 0.489 0.496 0.429 0.536 0.032 18 PLOS ONE | April 4, 2024 9 / 20 PLOS ONE Measuring fair income inequality in Thailand Fig 2. Scatter plots of the Cartesian coordinates of income Gini index and quintile income shares (QI Thailand from 1988 to 2021. (A) Bottom 20% (QI (QI between 0.00 and 1.00. i , i = 1, 2, 3, 4, 5) of 1). (B) Second 20% (QI 3 is between 0.00 and 0.25. The scale of QI 2). (C) Third 20% (QI 4). (E) Top 20% 4 is between 0.00 and 0.30 while the scale of QI 5 is 3). (D) Fourth 20% (QI 5). Note that the scale of QI 1; QI 2, and QI PLOS ONE | April 4, 2024 10 / 20
id: 5adeaa26b7459a4d66cd4754b0b9e787 - page: 9
PLOS ONE Measuring fair income inequality in Thailand The overall results indicate that the ordered pairs (income Gini index, QI i) are either above or below the fairness benchmarks in all five quintiles, with those in the bottom, the 2nd, and the top quintile being above their corresponding fairness benchmarks and those in the 3rd and the 4th quintile being below their corresponding fairness benchmarks. These results indicate that, relative to the fairness benchmarks, the Thai income earners in the bottom, the 2nd, and the top quintile receive income shares higher than the fair shares whereas the Thai income earners in the 3rd and the 4th quintile receive income shares lower than the fair shares. Given that the degree of unfairness in income shares by quintile is measured by the differ-
id: 3943a3d17950eaebc4c0441fb6a401c1 - page: 11
041) whereas, for the Thai income earners in the 3rd and the 4th quintile, both of which receive income shares lower than the fair shares, the income earners in the 4th quintile receive income share lower than the fair share the most ( = 0.049).
id: e3db9f1cc115deceff57781025cdd6eb - page: 11
Considering the degree of unfair income shares in each quintile as shown in Fig 2 and in Table 2, the results from the 1st quintile, where the income share is held by the bottom 20% (Fig 2A), suggest that the income share the income earners receive in 1990 is higher than the fair share the most ( = 0.016) whereas the income share the income earners receive in 2013 is higher than the fair share the least ( = 0.008). For the 2nd quintile where income share is held by the second 20% (Fig 2B), the results indicate that, in 2011, the income earners receive income share more than the fair share the most ( = 0.011) while, in 2019, the income earners receive income share more than the fair share the least ( = 0.003). The results from the 3rd quintile, where income share is held by the third 20% (Fig 2C), show that, in 2021, the income earners receive income share lower than the fair share the most ( = 0.017) whereas, in 2006, the income earners receive income share lower than the fair share th
id: 33edb8462960f4f6b211d5054a04c248 - page: 11
How to Retrieve?
# Search

curl -X POST "https://search.dria.co/hnsw/search" \
-H "x-api-key: <YOUR_API_KEY>" \
-H "Content-Type: application/json" \
-d '{"rerank": true, "top_n": 10, "contract_id": "nfBGoX7wz9tQTbZuS65Y58ltfqQBuje4BiA4S0O8OaM", "query": "What is alexanDRIA library?"}'
        
# Query

curl -X POST "https://search.dria.co/hnsw/query" \
-H "x-api-key: <YOUR_API_KEY>" \
-H "Content-Type: application/json" \
-d '{"vector": [0.123, 0.5236], "top_n": 10, "contract_id": "nfBGoX7wz9tQTbZuS65Y58ltfqQBuje4BiA4S0O8OaM", "level": 2}'