AdaSpot: Spend Resolution Where It Matters for Precise Event Spotting

Artur Xarles; Sergio Escalera; Thomas B. Moeslund; Albert Clapés

cs.CVFeb 25, 2026

AdaSpot: Spend Resolution Where It Matters for Precise Event Spotting

Authors: Artur Xarles, Sergio Escalera, Thomas B. Moeslund, Albert Clapés

Organizations: Universitat de Barcelona, Barcelona, Spain · Computer Vision Center, Cerdanyola del Vall`es, Spain · Aalborg University, Aalborg, Denmark

Abstract

Precise Event Spotting aims to localize fast-paced actions or events in videos with high temporal precision, a key task for applications in sports analytics, robotics, and autonomous systems. Existing methods typically process all frames uniformly, overlooking the inherent spatio-temporal redundancy in video data. This leads to redundant computation on non-informative regions while limiting overall efficiency. To remain tractable, they often spatially downsample inputs, losing fine-grained details crucial for precise localization. To address these limitations, we propose AdaSpot, a simple yet effective framework that processes low-resolution videos to extract global task-relevant features while adaptively selecting the most informative region-of-interest in each frame for high-resolution processing. The selection is performed via an unsupervised, task-aware strategy that maintains spatio-temporal consistency across frames and avoids the training instability of learnable alternatives. This design preserves essential fine-grained visual cues with a marginal computational overhead compared to low-resolution-only baselines, while remaining far more efficient than uniform high-resolution processing. Experiments on standard PES benchmarks demonstrate that AdaSpot achieves state-of-the-art performance under strict evaluation metrics (\eg, $+3.96$ and $+2.26$ mAP $@0$ frames on Tennis and FineDiving), while also maintaining strong results under looser metrics. Code is available at: https://github.com/arturxe2/AdaSpot.

Figures & tables

Figure 1: Illustration of standard PES approaches: (a) high-resolution videos incur high computational cost, whereas (b) low-resolution videos reduce cost but lose fine-grained details crucial for precise temporal localization. In contrast, (c) AdaSpot captures global context from low-resolution videos and adaptively applies high-resolution processing to task-relevant regions, preserving fine-grained details efficiently.

Figure 2: Overview of our proposed method, AdaSpot . (a) The framework uses a low-resolution extractor to process low-resolution clips and generate global features $F_{l}$ and spatial maps $F_{s}$ . A RoI selector leverages $F_{s}$ to identify the most relevant region in each frame. The resulting RoI sequence is then processed by a high-resolution extractor to capture fine-grained features, $F_{h}$ . $F_{l}$ and $F_{h}$ are linearly projected, aggregated, and passed through a temporal modeler , before a prediction head produces per-frame classifications. (b) Details of the RoI selector: channel averaging generates saliency maps from $F_{s}$ , spatio-temporal smoothing reduces noise, and adaptive-scale RoI selection adjusts the RoI size to the saliency spread.

	Tennis			FineDiving			FineGym			F3Set			Cost
Model	$\delta=0$ f	$1$ f	$2$ f	$0$ f	$1$ f	$2$ f	$0$ f	$1$ f	$2$ f	$0$ f	$1$ f	$2$ f	P(M)	GFLOPs
E2E-Spot ${}_{200\text{MF}}$ [ 14 ]	69.78 †	97.01 †	97.68 †	25.00 †	69.01 †	86.24 †	17.50 †	53.44 †	63.73 †	–	–	–	4.49	23.13
E2E-Spot ${}_{800\text{MF}}$ [ 14 ]	70.04 †	97.31 †	97.86 †	19.47 †	65.49 †	83.83 †	17.90 †	55.05 †	66.06 †	–	–	–	12.70	84.93
UGLF [ 42 ]	–	–	–	–	70.00	87.70	–	50.20	67.80	–	–	–	–	–
T-DEED ${}_{200\text{MF}}$ [ 51 ]	56.00 †	96.91 †	97.84 †	21.33 †	71.07 †	86.87 †	17.32 †	52.99 †	63.69 †	–	–	–	16.42	21.97
T-DEED ${}_{800\text{MF}}$ [ 51 ]	58.43 †	97.34 †	97.97 †	19.63 †	69.37 †	85.50 †	18.35 †	53.97 †	64.99 †	–	–	–	64.26	86.34

Table 1: Comparison with state-of-the-art methods for PES on Tennis, FineDiving, FineGym, and F3Set datasets. Bold and underlined values indicate the best and second-best results, respectively. The number of parameters (in millions) and GFLOPs for each method are also reported. For AdaSpot, we show results for two variants AdaSpot s and AdaSpot b (using RegNetY-200MF and RegNetY-400MF as feature extractors, respectively), reporting the mean over three random seeds along with the standard deviation (

\pm

std ). † denotes results obtained by re-running inference with the provided checkpoints using SNMS with a window of two and 50% overlapping for a fair comparison.

	SN-BAS		Cost
Model	$\delta$ = $0.5$ s	$1$ s	P(M)	GFLOPs
E2E-Spot ${}_{200\text{MF}}$ [ 14 ]	51.46	55.11	4.49	40.78
E2E-Spot ${}_{800\text{MF}}$ [ 14 ]	54.49	58.65	12.70	150.02
T-DEED ${}_{200\text{MF}}$ [ 51 ]	45.43	48.41	12.31	39.58
T-DEED ${}_{800\text{MF}}$ [ 51 ]	49.39	53.11	46.22	151.31
Santra et al. [ 34 ]	51.07	55.13	6.84	82.51

Table 2: Comparison of AdaSpot with state-of-the-art methods for the ES setting on SN-BAS. Bold and underlined values indicate the best and second-best results. The number of parameters (in millions) and GFLOPs for each method are also reported. Results for E2E-Spot and T-DEED are obtained by integrating their model into our training pipeline. For Santra et al. [34] , we implement our own version of the ASTRM module following the paper specifications due to the lack of publicly available code. For AdaSpot, we show results for two variants, AdaSpot s and AdaSpot b , and report the mean over three random seeds with standard deviation (

\pm

std ).

		Tennis			SN-BAS
Experiment		$\delta=0$ f	$1$ f	$2$ f	$\delta=0.5$ s	$1$ s
AdaSpot		73.30 $\pm$ 0.5	96.90 $\pm$ 0.1	97.47 $\pm$ 0.1	53.02 $\pm$ 0.5	56.43 $\pm$ 0.3
(a)	Single-branch counterparts
	low-res branch only	71.18	96.73	97.42	47.98	51.69
	high-res branch only	71.91	96.62	97.26	52.13	55.95
(b)	Padding type

Table 3: Ablation study of AdaSpot components on Tennis and SN-BAS, evaluating the impact of single branches, padding types, smoothing methods, fixed versus adaptive RoI scales, and auxiliary supervision. We report the mean over three random seeds along with the standard deviation (

\pm

std ).

Panel of Figure 3 — Figure 3: Comparison of AdaSpot, a single-branch baseline, and redundancy-aware alternatives across multiple spatial resolutions on Tennis (left) and SN-BAS (right). Each point corresponds to a model configuration, with GFLOPs on the x-axis, mAP on the y-axis, and point size indicating the number of parameters. Models closer to the upper-left with smaller markers achieve better accuracy-efficiency trade-offs.

Figure 4: Qualitative visualization of the saliency maps and the corresponding RoIs selected by AdaSpot across all evaluated datasets: Tennis, FineDiving, FineGym, F3Set, and SN-BAS. In FineDiving and FineGym, events revolve around a main athlete, whereas in Tennis, F3Set, and SN-BAS, they revolve around the ball, which is marked with a star for clarity.

Appendix figures & tables13 assets

Supplementary material from the paper’s appendix.

Appendix

Figure 5: Illustration of the taxonomy of methods addressing spatio-temporal redundancy. We categorize approaches into architecture-based and input-based , and indicate whether each method handles spatial redundancy, temporal redundancy, or both.

Figure 6: Qualitative visualization of saliency maps and the resulting selected RoIs when zero-padding is applied on the Tennis dataset. With zero-padding, the RoI selector ends up biased towards the central part of the frames.

		Tennis			SN-BAS
Experiment		$\delta=0$ f	$1$ f	$2$ f	$\delta=0.5$ s	$1$ s
AdaSpot		73.30	96.90	97.47	53.02	56.43
(a)	Fusion mechanism
	mean	71.26	96.48	97.07	50.55	54.56
	product	71.88	96.75	97.37	51.24	54.95
	linear	71.36	96.37	96.96	51.30	55.39

Appendix

Table 4: Extended ablation study of AdaSpot components on Tennis and SN-BAS, evaluating the impact of alternative fusion mechanisms, crop sizes, backbone reuse for both low- and high-resolution branches, adaptive RoI aspect ratios, and multiple RoIs per frame.

Figure 7: Sensitivity of AdaSpot performance ( $\text{mAP}@0\text{f}$ / $\text{mAP}@0.5\text{s}$ ) to variations in the threshold parameter $\tau$ . The blue line denotes the Tennis dataset (left y-axis), while the green line denotes the SN-BAS dataset (right y-axis).

	Low & high-res features			High-res features only
Dataset	AF-v2	Uni-AF	AdaSpot	AF-v2	Uni-AF	AdaSpot
Tennis	70.6 $\pm$ 1.5	70.2 $\pm$ 1.0	73.3 $\pm$ 0.5	68.8 $\pm$ 0.9	65.9 $\pm$ 0.9	71.9 $\pm$ 0.4
SN-BAS	49.0 $\pm$ 2.0	49.6 $\pm$ 1.3	53.0 $\pm$ 0.5	23.1 $\pm$ 15.1	39.4 $\pm$ 3.3	52.1 $\pm$ 0.7

Appendix

Table 5: Comparison of instability under the strictest metric for different learnable cropping methods with AdaSpot. Bold indicates best; (mean

\pm

std. across 3 runs).

Figure 8: Qualitative comparison of the RoIs selected by AdaSpot, AdaFocus-v2, and Uni-AdaFocus on the Tennis (left) and SN-BAS (right) datasets. For visualization, we mark the ball position in each frame with a star, as actions in these datsets occur around the ball; thus, relevant RoIs should contain or closely surround it.

	PES		ES
Model	P(M)	GFLOPs	P(M)	GFLOPs
E2E-Spot ${}_{200\text{MF}}$ [ 14 ]	4.49	23.13	4.49	40.78
E2E-Spot ${}_{800\text{MF}}$ [ 14 ]	12.70	84.93	12.70	150.02
T-DEED ${}_{200\text{MF}}$ [ 51 ]	16.42	21.97	12.31	39.58
T-DEED ${}_{800\text{MF}}$ [ 51 ]	64.26	86.34	46.22	151.31
Santra et al. [ 34 ]	6.46	57.84	6.84	82.51

Appendix

Table 6: Efficiency comparison of AdaSpot with state-of-the-art methods in both the PES setting (typically using

224{\times}224

inputs) and the ES setting (using

398{\times}224

inputs). For each configuration, we report the number of parameters (in millions) and the computational cost in GFLOPs.

		Tennis
Post-processing		$\delta=0$ f	$1$ f	$2$ f
NMS [ 30 ]	$\omega=1$	62.82	96.93	97.61
	$\omega=2$	62.45	96.61	97.64
	$\omega=3$	62.35	96.11	97.58
	$\omega=4$	62.32	95.94	97.47
	$\omega=5$	62.29	95.84	97.29

Appendix

Table 7: Post-processing sensitivity analysis on the Tennis dataset. We report results for standard NMS and Soft-NMS using different window sizes

\omega

. Bold and underlined values indicate the best and second-best results.

		AP ( $\delta=0$ f)
Event	Nº observations	E2E-Spot	T-DEED	AdaSpot	HR-only
Far-court ball bounce	8150	76.91	59.47	77.20	75.38
Near-court ball bounce	8127	76.79	68.19	78.98	78.91
Far-court swing	7123	53.24	41.52	64.76	58.77
Near-court swing	7044	56.42	48.85	58.83	58.49
Near-court serve	1690	76.79	67.91	79.24	78.67

Appendix

Table 8: Per-class analysis on the Tennis dataset. For each event class, we report the total number of observations and the AP

@0

f results for the best-performing versions of E2E-Spot, T-DEED, and AdaSpot, as well as for an AdaSpot variant using high-resolution features only (HR-only). Event classes are sorted in descending order of observations. The best result per class is highlighted in bold, and the second-best is underlined.

		AP ( $\delta=0$ f)
Event	Nº observations	E2E-Spot	T-DEED	AdaSpot
Entry	2984	22.51	24.02	26.74
Som(s).Pike	2152	27.21	23.14	27.58
Som(s).Tuck	1071	31.70	21.72	32.23
Twist(s)	803	18.60	16.45	22.51

Appendix

Table 9: Per-class analysis on the FineDiving dataset. For each event class, we report the total number of observations and the AP

@0

f results for the best-performing versions of E2E-Spot, T-DEED, and AdaSpot. Event classes are sorted in descending order of observations. The best result per class is highlighted in bold, and the second-best is underlined.

Event	Nº observations	E2E-Spot	T-DEED	AdaSpot
		AP ( $\delta=0$ f)
Uneven bars circles start	6612	11.32	10.26	10.28
Uneven bars circles end	6612	20.19	19.89	19.63
Balance beam leap_jump_hop start	4787	17.52	19.72	18.07
Balance beam leap_jump_hop end	4787	10.31	12.64	10.81
Balance beam flight_salto start	4187	19.84	22.72	24.35

Appendix

Table 10: Per-class analysis on the FineGym dataset. For each event class, we report the total number of observations and the AP

@0

		AP ( $\delta=0.5$ s)
Event	Nº observations	E2E-Spot	T-DEED	AdaSpot	HR-only
Pass	4985	85.15	83.44	85.94	85.11
Drive	4300	81.55	77.25	81.77	81.50
High pass	761	79.30	76.33	78.68	75.18
Header	713	68.14	54.27	68.87	62.62
Ball out of play	551	16.97	19.79	23.01	21.15

Appendix

Table 11: Per-class analysis on the SN-BAS dataset. For each event class, we report the total number of observations and the AP

@0.5

s results for the best-performing versions of E2E-Spot, T-DEED, and AdaSpot, as well as for an AdaSpot variant using high-resolution features only (HR-only). Event classes are sorted in descending order of observations. The best result per class is highlighted in bold, and the second-best is underlined.

F 3 ED [ 24 ]	24.8	60.7	64.8	40.3	74.0
	mAP $@0$ f	mAP $@1$ f	mAP $@2$ f	F1 evt	Edit
AdaSpot s	53.55 $\pm$ 1.2	67.76 $\pm$ 0.8	68.41 $\pm$ 1.0	48.8 $\pm$ 1.1	72.6 $\pm$ 0.4
AdaSpot b	55.38 $\pm$ 0.3	69.37 $\pm$ 0.2	69.94 $\pm$ 0.2	51.66 $\pm$ 0.6	73.66 $\pm$ 0.4

Appendix

Table 12: Comparison of AdaSpot with F 3 ED on the F3Set dataset using standard PES mAP metrics, as well as F1 and Edit scores. Results show the mean over three random seeds with the corresponding standard deviation (

\pm

) . Bold and underlined values indicate the best and second-best results.

Temporal Action Localization

Temporal action localization finds the start and end times of actions in untrimmed videos.

See all 32 papers

Temporal Feature Distillation for Label-Efficient Precise Event Spotting in Sports Videos
Jul 13, 2026Hao Xu, Xinyu Wei, Sam Wells +1Vision Transformer Sports Video Analysis
Deakin University Melbourne, Australia · Champion Data Melbourne, Australia · Paralympics Australia Melbourne, Australia
ScanFocus: A Coarse-to-Fine Framework for Spatio-Temporal Video Grounding
Jul 15, 2026Kai Chen, Ming Dai, Wenxuan Cheng +1Temporal Video Grounding Vision-Language Grounding
Southeast University
TubeLite: Lightweight Multi-Actor Spatio-Temporal Action Detection
Jul 6, 2026Ali Soltaninezhad, Melissa Cote, Alejandro Rico Espinosa +2Temporal Action Detection Temporal Video Understanding
Electrical and Computer Engineering, University of Victoria, Victoria, BC, Canada

Jul 13, 2026J/K move · Enter open · S save

Temporal Feature Distillation for Label-Efficient Precise Event Spotting in Sports Videos

Hao Xu, Xinyu Wei, Sam Wells, Sunil Aryal

Deakin University Melbourne, Australia · Champion Data Melbourne, Australia · Paralympics Australia Melbourne, Australia

Vision Transformer Sports Video Analysis Semi-Supervised Learning Temporal Action Detection

Open paper AskPDF arXiv

Precise Event Spotting (PES) requires distinguishing visually similar yet semantically distinct adjacent frames, making it fundamentally different from image classification and coarse action recognition. Although self-distillation methods such as DINO have shown strong representation learning ability in images, we find that directly applying them to PES is ineffective: without supervised guidance, subtle but crucial motion cues are often suppressed as noise, leading to representations that are insensitive to precise event boundaries. To address this, we propose Temporal Feature Distillation, a semi-supervised objective that aligns temporally informative backbone features, rather than projection-head outputs, to preserve motion-sensitive and boundary-aware cues for frame-level localization. A supervised warm-up with a ramp-up schedule further stabilizes training by ensuring that meaningful event cues are learned before unlabeled distillation begins. We also introduce Transformer Gate Shift, a multi-scale gated shifting module that injects motion-aware temporal information into Vision Transformers. Experiments on four fine-grained sports benchmarks show consistent improvements over fully supervised and semi-supervised baselines. Under 10% supervision on FSPerf, our method improves mAP by 4.54 points over the strongest competing approach, and with only 80% labeled data, it matches or surpasses the fully supervised 100% baseline on two of the four datasets.

Abstract

Temporal Action Localization

Explore similar work

Temporal Feature Distillation for Label-Efficient Precise Event Spotting in Sports Videos

ScanFocus: A Coarse-to-Fine Framework for Spatio-Temporal Video Grounding

TubeLite: Lightweight Multi-Actor Spatio-Temporal Action Detection