Open-source Vehicle Detection Model - Supports Multi-type Vehicle Recognition and Localization

Home

Vehicle Detection Model

Developed by ArrayDice

Vehicle detection model based on DETR framework, supporting recognition and localization of multiple vehicle types

Object Detection

Transformers

Open Source License:Apache-2.0 #Multi-object detection #Traffic scenarios #DETR architecture

Downloads 32

Release Time : 7/11/2024

Model Overview

This model is a fine-tuned vehicle detection model based on facebook/detr-resnet-50, capable of recognizing various vehicle types such as RVs, sedans, pickup trucks, and providing their location information in images.

Model Features

End-to-end object detection

Adopts DETR architecture, eliminating the need for anchor box design in traditional object detection, achieving an end-to-end detection process

Multi-category vehicle recognition

Capable of recognizing various vehicle types such as RVs, sedans, and pickup trucks

Detection at different scales

Capable of detecting vehicle targets of small, medium, and large sizes

Model Capabilities

Vehicle detection

Object localization

Multi-category recognition

Use Cases

Intelligent transportation

Traffic monitoring analysis

Used for vehicle detection and statistics in traffic surveillance cameras

Can identify different types of vehicles and count their numbers

Autonomous driving perception

Provides surrounding vehicle detection capability for autonomous driving systems

Helps autonomous vehicles recognize the position and type of surrounding vehicles

Security surveillance

Parking lot management

Used for vehicle entry/exit detection and classification in parking lots

Can track parking status of different vehicle types

🚀 Vehicle_Detection_Model

This Vehicle Detection Model is a fine - tuned version of [facebook/detr - resnet - 50](https://huggingface.co/facebook/detr - resnet - 50). It aims to accurately detect vehicles in relevant scenarios, providing reliable results on the evaluation set.

🚀 Quick Start

This model is a fine - tuned version of [facebook/detr - resnet - 50](https://huggingface.co/facebook/detr - resnet - 50) on the None dataset. It achieves the following results on the evaluation set:

Loss: 0.7220
Map: 0.0875
Map 50: 0.1634
Map 75: 0.084
Map Small: 0.355
Map Medium: 0.1423
Map Large: 0.0499
Mar 1: 0.1462
Mar 10: 0.2602
Mar 100: 0.2709
Mar Small: 0.5
Mar Medium: 0.4013
Mar Large: 0.3
Map Camping car: 0.0039
Mar 100 Camping car: 0.35
Map Car: 0.4971
Mar 100 Car: 0.6256
Map Other: 0.0
Mar 100 Other: 0.0
Map Pickup: 0.0239
Mar 100 Pickup: 0.65
Map Truck: 0.0
Mar 100 Truck: 0.0
Map Van: 0.0
Mar 100 Van: 0.0

📚 Documentation

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

learning_rate: 5e - 05
train_batch_size: 8
eval_batch_size: 8
seed: 42
optimizer: Adam with betas=(0.9,0.999) and epsilon = 1e - 08
lr_scheduler_type: cosine
num_epochs: 30

Training results

Training Loss	Epoch	Step	Validation Loss	Map	Map 50	Map 75	Map Small	Map Medium	Map Large	Mar 1	Mar 10	Mar 100	Mar Small	Mar Medium	Mar Large	Map Camping car	Mar 100 Camping car	Map Car	Mar 100 Car	Map Pickup	Mar 100 Pickup
No log	1.0	232	1.3186	0.0125	0.0292	0.0085	0.0059	0.0197	0.004	0.0182	0.0467	0.0903	0.0556	0.1396	0.1	0.0	0.0	0.0751	0.5416	0.0	0.0
No log	2.0	464	1.0802	0.0375	0.0832	0.0283	0.263	0.0585	0.0102	0.0254	0.0734	0.0865	0.2667	0.1317	0.15	0.0	0.0	0.2249	0.5189	0.0	0.0
1.4746	3.0	696	0.9804	0.0608	0.125	0.0483	0.2948	0.0938	0.075	0.0305	0.0821	0.0903	0.3556	0.1372	0.075	0.0	0.0	0.3646	0.5416	0.0	0.0
1.4746	4.0	928	0.9139	0.0712	0.1426	0.0577	0.2232	0.1091	0.125	0.0315	0.085	0.0959	0.3333	0.1459	0.125	0.0	0.0	0.4274	0.5754	0.0	0.0
1.0134	5.0	1160	0.9581	0.0637	0.1367	0.0426	0.0804	0.0993	0.075	0.0289	0.0771	0.0846	0.2556	0.1292	0.075	0.0	0.0	0.3823	0.5078	0.0	0.0
1.0134	6.0	1392	0.8830	0.0743	0.1476	0.0645	0.2171	0.1139	0.075	0.0345	0.0862	0.0967	0.3222	0.1474	0.075	0.0	0.0	0.4456	0.5801	0.0	0.0
0.9133	7.0	1624	0.8645	0.0716	0.147	0.0571	0.2156	0.1099	0.0752	0.0329	0.0853	0.0966	0.3111	0.147	0.175	0.0	0.0	0.4296	0.5797	0.0	0.0
0.9133	8.0	1856	0.8776	0.0676	0.1478	0.0445	0.2225	0.1032	0.1254	0.0317	0.0811	0.0934	0.3333	0.1417	0.2	0.0	0.0	0.4056	0.5601	0.0	0.0
0.8617	9.0	2088	0.8556	0.0754	0.1506	0.0638	0.252	0.1153	0.1254	0.0323	0.0881	0.0999	0.3667	0.1517	0.2	0.0	0.0	0.4525	0.5996	0.0	0.0
0.8617	10.0	2320	0.7552	0.0821	0.1528	0.0803	0.361	0.1248	0.1257	0.0347	0.1119	0.1312	0.5222	0.1747	0.2625	0.0	0.0	0.4921	0.6206	0.0004	0.1667
0.8182	11.0	2552	0.8211	0.0741	0.1525	0.0627	0.3014	0.114	0.0503	0.0381	0.1098	0.1544	0.4222	0.2641	0.125	0.0001	0.1	0.4422	0.5762	0.002	0.25
0.8182	12.0	2784	0.8140	0.0734	0.1486	0.056	0.214	0.1129	0.0506	0.0451	0.1066	0.1184	0.3778	0.1941	0.125	0.0	0.0	0.4382	0.594	0.0024	0.1167
0.8079	13.0	3016	0.7473	0.0848	0.1559	0.0789	0.2794	0.1307	0.0779	0.078	0.16	0.1727	0.5	0.2668	0.1625	0.0	0.0	0.4983	0.6363	0.0103	0.4
0.8079	14.0	3248	0.8514	0.0706	0.1541	0.0485	0.3034	0.1071	0.1046	0.079	0.1726	0.1862	0.4222	0.2556	0.2875	0.0	0.0	0.4149	0.5669	0.0089	0.55
0.8079	15.0	3480	0.7615	0.0814	0.1579	0.0709	0.3673	0.1229	0.0691	0.0929	0.161	0.1719	0.4444	0.2177	0.3125	0.0	0.0	0.4774	0.6146	0.0109	0.4167
0.7672	16.0	3712	0.7819	0.0786	0.1555	0.0732	0.2782	0.1182	0.1589	0.103	0.1757	0.1878	0.3556	0.2629	0.325	0.0	0.0	0.4557	0.5936	0.0159	0.5333
0.7672	17.0	3944	0.7723	0.0807	0.1551	0.0757	0.3302	0.1227	0.0687	0.0812	0.1762	0.1939	0.4	0.274	0.3	0.0	0.0	0.4724	0.6135	0.0121	0.55
0.744	18.0	4176	0.7838	0.0784	0.1535	0.0709	0.3594	0.1201	0.0564	0.0865	0.1865	0.2497	0.4	0.3637	0.325	0.0009	0.3	0.4503	0.5982	0.0191	0.6
0.744	19.0	4408	0.7714	0.0763	0.1552	0.0625	0.2732	0.116	0.0675	0.0605	0.1696	0.1824	0.4111	0.2502	0.3	0.0	0.0	0.4461	0.5943	0.0115	0.5
0.7219	20.0	4640	0.7403	0.0808	0.1543	0.0686	0.2968	0.1251	0.0665	0.0984	0.1889	0.1997	0.5111	0.2859	0.3	0.0	0.0	0.4738	0.6146	0.0109	0.5833
0.7219	21.0	4872	0.7421	0.0839	0.1599	0.0803	0.2862	0.2166	0.0451	0.1396	0.2437	0.257	0.5	0.3782	0.275	0.0123	0.4	0.4792	0.6089	0.0118	0.5333
0.6955	22.0	5104	0.7459	0.0844	0.1619	0.0753	0.3116	0.1475	0.0462	0.1395	0.2253	0.2342	0.4333	0.3404	0.2875	0.0068	0.25	0.4815	0.6053	0.018	0.55
0.6955	23.0	5336	0.7439	0.0844	0.1594	0.0782	0.3246	0.1335	0.0682	0.1559	0.2468	0.2567	0.4667	0.378	0.3	0.0031	0.3	0.4854	0.6071	0.018	0.6333
0.6836	24.0	5568	0.7263	0.0879	0.1619	0.0854	0.3509	0.1397	0.0597	0.1718	0.2499	0.2589	0.4333	0.3818	0.275	0.0027	0.3	0.4998	0.6203	0.025	0.6333
0.6836	25.0	5800	0.7146	0.0881	0.1632	0.087	0.3606	0.1412	0.0714	0.1518	0.2466	0.2572	0.4667	0.3768	0.3	0.0032	0.3	0.4988	0.6263	0.0267	0.6167
0.6671	26.0	6032	0.7180	0.0915	0.1678	0.0918	0.3652	0.2262	0.0483	0.163	0.2506	0.2672	0.5	0.3979	0.2875	0.0186	0.35	0.5069	0.6367	0.0235	0.6167
0.6671	27.0	6264	0.7219	0.0907	0.1678	0.0888	0.3596	0.2122	0.0492	0.1375	0.26	0.2709	0.5	0.4012	0.3	0.0184	0.35	0.5014	0.6253	0.0243	0.65
0.6671	28.0	6496	0.7177	0.0882	0.1642	0.0866	0.3625	0.1439	0.0505	0.1463	0.2603	0.2715	0.5111	0.4021	0.3	0.0038	0.35	0.4996	0.6288	0.0256	0.65
0.656	29.0	6728	0.7257	0.0873	0.1634	0.084	0.355	0.1417	0.0497	0.1433	0.2572	0.2708	0.5	0.401	0.3	0.0039	0.35	0.4962	0.6246	0.0237	0.65
0.656	30.0	6960	0.7220	0.0875	0.1634	0.084	0.355	0.1423	0.0499	0.1462	0.2602	0.2709	0.5	0.4013	0.3	0.0039	0.35	0.4971	0.6256	0.0239	0.65

Framework versions

Transformers 4.41.2
Pytorch 2.3.0+cu121
Datasets 2.20.0
Tokenizers 0.19.1

📄 License

This model is licensed under the Apache 2.0 license.

Featured Recommended AI Models

Empowering the Future, Your AI Solution Knowledge Base

English 简体中文繁體中文にほんご