（武汉9日综合电）中国武汉大学计算机学院日前研发了一款隐身外套（InvisDefense），不论是白天或夜晚，只要穿上这件外套，就可以躲过中共设置的“智能人脸辨识”监视器；就算被捕捉到，也能“人为欺骗”监视器，让它无法判断你是不是人类。
武汉大学计算机学院于11月底公布这个作品，并在中国研究生创新实践大赛创意作品中获得第一名。这个研究项目由王政教授主持，这件外套外观朴素且成本低廉，可以将人体隐藏在AI监控的摄影机之外。
研究团队称，虽然外套肉眼可见，但覆盖在外套上的图案白天会遮蔽摄影机镜头，晚上则是会发出异常的热信号。王政教授表示，因为中国道路上的监视摄影机可以辨别行人、道路和障碍物，只要穿上这件外套，马上可以达到“隐身”目的。
据悉，这件外套能够在白天干扰摄影机的演算法，使镜头“失明”，进而无法分辨穿戴隐身外套者的真实身分。晚上，由于摄影机会根据红外线热影像跟踪人体，外套表层的不规则形状控制模组会产生不寻常的温度模式，藉此混淆监视器。
王政教授指出，经过目前检测，隐身外套可降低57%监视器人脸监测功能，他预估这个数字未来将会更高。
SQL: Array
(
[0] => Array
(
[query] => SELECT * FROM "wp_posts" WHERE "wp_posts"."ID" = :MeD0_mK AND "wp_posts"."post_status" = :MeD1_mK LIMIT 1
[params] =>
[time_ms] => 0.473022
)
[1] => Array
(
[query] => SELECT tt.term_id as tagid from wp_term_relationships tr
left join `wp_term_taxonomy` tt On tt.term_taxonomy_id = tr.term_taxonomy_id
WHERE tt.taxonomy IN ('post_tag')
AND tr.object_id = 3232923
[params] =>
[time_ms] => 0.448942
)
[2] => Array
(
[query] => SELECT "wp_terms"."term_id" AS "term_id","wp_terms"."name" AS "name","wp_terms"."slug" AS "slug","wp_terms"."term_order" AS "term_order","wp_term_taxonomy"."description" AS "description","wp_term_taxonomy"."parent" AS "parent","wp_term_taxonomy"."count" AS "count","wp_term_taxonomy"."taxonomy" AS "tax" FROM "wp_terms" LEFT JOIN "wp_term_taxonomy" USING ("term_id") WHERE "wp_terms"."term_id" = :MeD2_mK AND "wp_term_taxonomy"."taxonomy" = :MeD3_mK LIMIT 1
[params] =>
[time_ms] => 0.428915
)
[3] => Array
(
[query] => SELECT "wp_terms"."term_id" AS "term_id","wp_terms"."name" AS "name","wp_terms"."slug" AS "slug","wp_terms"."term_order" AS "term_order","wp_term_taxonomy"."description" AS "description","wp_term_taxonomy"."parent" AS "parent","wp_term_taxonomy"."count" AS "count","wp_term_taxonomy"."taxonomy" AS "tax" FROM "wp_terms" LEFT JOIN "wp_term_taxonomy" USING ("term_id") WHERE "wp_terms"."term_id" = :MeD4_mK AND "wp_term_taxonomy"."taxonomy" = :MeD5_mK LIMIT 1
[params] =>
[time_ms] => 0.358105
)
[4] => Array
(
[query] => SELECT "meta_key","meta_value" FROM "wp_postmeta" WHERE "post_id" = :MeD6_mK
[params] =>
[time_ms] => 0.422955
)
[5] => Array
(
[query] => SELECT * FROM "wp_posts" WHERE "wp_posts"."ID" = :MeD7_mK AND "wp_posts"."post_status" = :MeD8_mK LIMIT 1
[params] =>
[time_ms] => 0.356913
)
[6] => Array
(
[query] => SELECT * FROM "wp_posts" WHERE "wp_posts"."ID" = :MeD9_mK AND "wp_posts"."post_status" = :MeD10_mK LIMIT 1
[params] =>
[time_ms] => 0.250816
)
[7] => Array
(
[query] => SELECT * FROM "wp_posts" WHERE "wp_posts"."ID" = :MeD11_mK AND "wp_posts"."post_status" = :MeD12_mK LIMIT 1
[params] =>
[time_ms] => 0.247955
)
[8] => Array
(
[query] => SELECT tt.term_id as tagid from wp_term_relationships tr
left join `wp_term_taxonomy` tt On tt.term_taxonomy_id = tr.term_taxonomy_id
WHERE tt.taxonomy IN ('post_tag')
AND tr.object_id = 3232923
[params] =>
[time_ms] => 0.247955
)
[9] => Array
(
[query] => SELECT "wp_terms"."term_id" AS "term_id","wp_terms"."name" AS "name","wp_terms"."slug" AS "slug","wp_terms"."term_order" AS "term_order","wp_term_taxonomy"."description" AS "description","wp_term_taxonomy"."parent" AS "parent","wp_term_taxonomy"."count" AS "count","wp_term_taxonomy"."taxonomy" AS "tax" FROM "wp_terms" LEFT JOIN "wp_term_taxonomy" USING ("term_id") WHERE "wp_terms"."term_id" = :MeD13_mK AND "wp_term_taxonomy"."taxonomy" = :MeD14_mK LIMIT 1
[params] =>
[time_ms] => 0.195026
)
[10] => Array
(
[query] => SELECT "wp_terms"."term_id" AS "term_id","wp_terms"."name" AS "name","wp_terms"."slug" AS "slug","wp_terms"."term_order" AS "term_order","wp_term_taxonomy"."description" AS "description","wp_term_taxonomy"."parent" AS "parent","wp_term_taxonomy"."count" AS "count","wp_term_taxonomy"."taxonomy" AS "tax" FROM "wp_terms" LEFT JOIN "wp_term_taxonomy" USING ("term_id") WHERE "wp_terms"."term_id" = :MeD15_mK AND "wp_term_taxonomy"."taxonomy" = :MeD16_mK LIMIT 1
[params] =>
[time_ms] => 0.192881
)
)
Redis: Array
(
[0] => Array
(
[command] => RAW
[arguments] => Array
(
[0] => HGET
[1] => post_info:3232923
[2] => info
)
[time_ms] => 0.2
)
[1] => Array
(
[command] => RAW
[arguments] => Array
(
[0] => HGET
[1] => category_info:all
[2] => info
)
[time_ms] => 0.58
)
[2] => Array
(
[command] => RAW
[arguments] => Array
(
[0] => ZRANGE
[1] => termlist27872
[2] => 0
[3] => 7
)
[time_ms] => 0.19
)
[3] => Array
(
[command] => RAW
[arguments] => Array
(
[0] => HGET
[1] => post_info:4589333
[2] => info
)
[time_ms] => 0.2
)
[4] => Array
(
[command] => RAW
[arguments] => Array
(
[0] => HGET
[1] => post_info:3029082
[2] => info
)
[time_ms] => 0.15
)
[5] => Array
(
[command] => RAW
[arguments] => Array
(
[0] => HGET
[1] => post_info:1644325
[2] => info
)
[time_ms] => 0.17
)
[6] => Array
(
[command] => RAW
[arguments] => Array
(
[0] => ZRANGE
[1] => termlist120029
[2] => 0
[3] => 7
)
[time_ms] => 0.18
)
[7] => Array
(
[command] => RAW
[arguments] => Array
(
[0] => HGET
[1] => post_info:4979603
[2] => info
)
[time_ms] => 0.2
)
[8] => Array
(
[command] => RAW
[arguments] => Array
(
[0] => HGET
[1] => post_info:4947909
[2] => info
)
[time_ms] => 0.17
)
[9] => Array
(
[command] => RAW
[arguments] => Array
(
[0] => HGET
[1] => post_info:4876807
[2] => info
)
[time_ms] => 0.22
)
[10] => Array
(
[command] => RAW
[arguments] => Array
(
[0] => HGET
[1] => post_info:4705991
[2] => info
)
[time_ms] => 0.2
)
[11] => Array
(
[command] => RAW
[arguments] => Array
(
[0] => HGET
[1] => post_info:4542085
[2] => info
)
[time_ms] => 0.18
)
[12] => Array
(
[command] => RAW
[arguments] => Array
(
[0] => HGET
[1] => post_info:4430438
[2] => info
)
[time_ms] => 0.17
)
[13] => Array
(
[command] => RAW
[arguments] => Array
(
[0] => HGET
[1] => post_info:4139773
[2] => info
)
[time_ms] => 0.19
)
)
