»ùÓÚÊÓ¾õµÄ×Ô¶¯¼ÝʻϵͳÐèÒª»ùÓÚµ¥Ä¿ÉãÏñÍ·»ñÈ¡µÄͼÏñ£¬£¬£¬£¬£¬£¬ÅжÏÄ¿½ñ³µÁ¾ÓëÖÜΧ³µÁ¾¡¢ÐÐÈ˺ÍÕÏ°­ÎïµÄ¾àÀ룬£¬£¬£¬£¬£¬¾àÀëÅжϵľ«¶È¶Ô×Ô¶¯¼ÝʻϵͳµÄÇå¾²ÐÔÓÐמöÒéÐÔµÄÓ°Ï죬£¬£¬£¬£¬£¬ÉÌÌÀ¿Æ¼¼ÔÚCVPR 2018½ÒÏþÁÁµã±¨¸æ(Spotlight)ÂÛÎÄ£¬£¬£¬£¬£¬£¬Ìá³ö»ùÓÚµ¥Ä¿Í¼ÏñµÄÉî¶ÈÔ¤ÅÌËã·¨£¬£¬£¬£¬£¬£¬´ó·ù¶ÈÌáÉý»ùÓÚµ¥Ä¿Í¼ÏñÉî¶ÈÔ¤¼ÆµÄ¾«¶È£¬£¬£¬£¬£¬£¬½øÒ»²½ÌáÉý×Ô¶¯¼ÝʻϵͳµÄÇå¾²ÐÔ¡£¡£¡£¡£¡£¡£¡£¸ÃÂÛÎÄÓÉÉÌÌÀ¿Æ¼¼¼ûÏ°Ñо¿Ô±ÂÞÔ½ÔÚÑо¿ÔºÑо¿Ô±ÈÎ˼½ÝÖ¸µ¼ÏÂÍê³É¡£¡£¡£¡£¡£¡£¡£±¾ÎÄΪÉÌÌÀ¿Æ¼¼CVPR 2018ÂÛÎĽâ¶ÁµÚ5ÆÚ¡£¡£¡£¡£¡£¡£¡£

ÂÛÎÄ£ºSingle View Stereo Matching

×÷ÕߣºYue Luo, Jimmy Ren, Mude Lin, Jiahao Pang, Wenxiu Sun, Hongsheng Li, Liang Lin

ÂÛÎĵص㣺

https://arxiv.org/abs/1803.02612

´úÂëµØµã£º

https://github.com/lawy623/SVS

¼ò½é

»ùÓÚµ¥Ä¿Í¼ÏñµÄÉî¶ÈÔ¤ÅÌËã·¨¾ßÓÐÀû±ã°²ÅÅ¡¢ÅÌË㱾ǮµÍµÈÓŵ㣬£¬£¬£¬£¬£¬Êܵ½ÁËѧÊõ½çºÍ¹¤Òµ½çÈÕÒæÔöÌíµÄ¹Ø×¢¡£¡£¡£¡£¡£¡£¡£ÏÖÓеĵ¥Ä¿Éî¶ÈÔ¤¼ÆÒªÁìͨ³£Ê¹Óüòµ¥ÊӽǵÄͼÏñÊý¾Ý×÷ΪÊäÈ룬£¬£¬£¬£¬£¬Ö±½ÓÕ¹ÍûͼÏñÖÐÿ¸öÏñËضÔÓ¦µÄÉî¶ÈÖµ£¬£¬£¬£¬£¬£¬ÕâÖÖ½â¾ö¼Æ»®µ¼ÖÂÏÖÓÐÒªÁìͨ³£ÐèÒª´ó×ÚµÄÉî¶È±ê×¢Êý¾Ý£¬£¬£¬£¬£¬£¬¶øÕâÀàÊý¾Ýͨ³£ÐèÒª½Ï¸ßµÄÊÕÂÞ±¾Ç®¡£¡£¡£¡£¡£¡£¡£½üÄêÀ´µÄË¢ÐÂ˼Ð÷Ö÷ÒªÊÇÔÚѵÁ·Àú³ÌÖÐÒýÈëÒþʽµÄ¼¸ºÎÔ¼Êø£¬£¬£¬£¬£¬£¬Í¨¹ý¼¸ºÎ±ä»»£¬£¬£¬£¬£¬£¬Ê¹ÓÃÒ»²àÉãÏñ»úͼÏñ(ÒÔϳÆÓÒͼ)¼àÊÓ»ùÓÚÁíÒ»²àÉãÏñ»úͼÏñ(ÒÔϳÆ×óͼ)Õ¹ÍûµÄÉî¶Èͼ£¬£¬£¬£¬£¬£¬´Ó¶øïÔÌ­¶ÔÊý¾ÝµÄÒÀÀµ¡£¡£¡£¡£¡£¡£¡£µ«ÕâÀàÒªÁìÔÚ²âÊÔÀú³ÌÖÐÈÔȻȱ·¦ÏÔʽµÄ¼¸ºÎÔ¼Êø¡£¡£¡£¡£¡£¡£¡£ÎªÏàʶ¾öÉÏÊöÎÊÌ⣬£¬£¬£¬£¬£¬±¾ÎÄÌá³öµ¥ÊÓͼ˫ĿƥÅäÄ£×Ó(Single View Stereo Matching, SVS)£¬£¬£¬£¬£¬£¬¸ÃÄ£×Ӱѵ¥Ä¿Éî¶ÈÔ¤¼ÆÆÊÎöΪÁ½¸ö×ÓÀú³Ì£¬£¬£¬£¬£¬£¬ÊÓͼºÏ³ÉÀú³ÌºÍË«Ä¿Æ¥ÅäÀú³Ì£¬£¬£¬£¬£¬£¬ÆäËã·¨¿ò¼ÜÈçͼ1Ëùʾ¡£¡£¡£¡£¡£¡£¡£

ͼ1£ºµ¥ÊÓͼ˫ĿƥÅäÄ£×ÓµÄʾÒâͼ

ͨ¹ýÕâÑùµÄÆÊÎö£¬£¬£¬£¬£¬£¬Ê¹µÃÌá³öµÄÄ£×ÓÓÐÈçÏÂÁ½¸öÓŵ㣺

• ¼«´óµØïÔÌ­Éî¶È±ê×¢Êý¾ÝµÄÒÀÀµ;

• ÔÚ²âÊԽ׶ÎÏÔʽµØÒýÈ뼸ºÎÔ¼Êø¡£¡£¡£¡£¡£¡£¡£

ʵÑé֤ʵ£¬£¬£¬£¬£¬£¬±¾ÎÄÌá³öµÄÄ£×Ó½öÓÃÉÙÁ¿µÄÉî¶È±ê×¢Êý¾Ý¾Í¿ÉÒÔÔÚKITTIÊý¾Ý¼¯ÉÏÁè¼Ý֮ǰµÄËùÓе¥Ä¿Éî¶ÈÔ¤¼ÆÒªÁ죬£¬£¬£¬£¬£¬²¢Ê״νö¿¿µ¥Ä¿Í¼ÏñÊý¾Ý¾ÍÁè¼ÝÁËË«Ä¿Æ¥ÅäËã·¨Block MatchingµÄÉî¶ÈÔ¤¼Æ¾«¶È¡£¡£¡£¡£¡£¡£¡£

SVSÄ£×Ó

ÏÖÓлùÓÚÉî¶ÈѧϰµÄµ¥Ä¿Éî¶ÈÔ¤¼ÆÒªÁ죬£¬£¬£¬£¬£¬Í¨³£°ÑCNN×÷ΪºÚºÐʹÓ㬣¬£¬£¬£¬£¬Ñ§Ï°Í¼Ïñ¿éÖÁÉî¶ÈÖµµÄÖ±½ÓÓ³É䣬£¬£¬£¬£¬£¬ÕâÀàÒªÁìÍêÈ«ÒÀÀµ¸ß¼¶ÓïÒåÐÅÏ¢×÷ΪչÍûÉî¶ÈµÄÒÀ¾Ý£¬£¬£¬£¬£¬£¬Ö»¹ÜÓÐЩҪÁìÔÚËðʧº¯ÊýÉÏÒýÈëһЩÌØÊâµÄÔ¼ÊøÌõ¼þ£¬£¬£¬£¬£¬£¬Ñ§Ï°ÕâÑùµÄÓïÒåÐÅÏ¢ÈÔÈ»ÊǺÜÊÇÄÑÌâµÄ¡£¡£¡£¡£¡£¡£¡£ÁíÒ»·½Ã棬£¬£¬£¬£¬£¬×ÝÈ»ÕâÑùµÄÓ³ÉäÄܹ»±»ÀÖ³ÉѵÁ·£¬£¬£¬£¬£¬£¬Ë㷨ͨ³£Ò²ÐèÒª´ó×Ú´øÉî¶ÈÖµ±êÇ©µÄÕæʵÊý¾Ý£¬£¬£¬£¬£¬£¬¶øÕâÀàÊý¾ÝµÄÊÕÂÞ±¾Ç®ºÜÊǸßÇÒºÄʱ£¬£¬£¬£¬£¬£¬¼«´óµÄÏÞÖÆÁËÕâÀàÊÖÒÕµÄÊÊÓó¡¾°¡£¡£¡£¡£¡£¡£¡£

»ùÓÚÉÏÊöÆÊÎö£¬£¬£¬£¬£¬£¬±¾ÎÄÒªÁìÌá³öÁËÒ»ÖÖÐÂÓ±µÄÃæÏòµ¥Ä¿Éî¶ÈÔ¤¼ÆµÄËã·¨¿ò¼Ü£¬£¬£¬£¬£¬£¬°Ñµ¥Ä¿Éî¶ÈÔ¤¼ÆÆÊÎöΪÁ½¸öÀú³Ì£¬£¬£¬£¬£¬£¬¼´ÊÓͼºÏ³ÉÀú³ÌºÍË«Ä¿Æ¥ÅäÀú³Ì¡£¡£¡£¡£¡£¡£¡£Ä£×ÓµÄÖ÷ÒªÉè¼Æ˼Ð÷ÔÚÓÚ£º

• °ÑË«Ä¿Éî¶ÈÔ¤¼ÆÄ£×ÓÖÐÓÐÓõļ¸ºÎÔ¼ÊøÏÔʽµØÍŽᵽµ¥Ä¿Éî¶ÈÔ¤¼ÆÄ£×ÓÖУ¬£¬£¬£¬£¬£¬Ìá¸ßÄ£×ӵĿÉÚ¹ÊÍÐÔ;

• ïÔ̭ʹÓÃÄÑÒÔÊÕÂÞµÄÕæʵÉî¶ÈÊý¾Ý£¬£¬£¬£¬£¬£¬´Ó¶øÀ©´óÄ£×ÓµÄÊÊÓùæÄ£;

• Õû¸öÄ£×ÓÒԶ˵½¶ËµÄµÄ·½·¨ÑµÁ·£¬£¬£¬£¬£¬£¬´Ó¶øÌáÉýÉî¶ÈÔ¤¼Æ׼ȷÐÔ¡£¡£¡£¡£¡£¡£¡£

Ä£×ÓµÄÊÓͼºÏ³ÉÀú³ÌÓÉÊÓͼºÏ³ÉÍøÂçÍê³É£¬£¬£¬£¬£¬£¬ÊäÈëÒ»ÕÅ×óͼ£¬£¬£¬£¬£¬£¬ÍøÂçºÏ³É¸ÃͼÏñ¶ÔÓ¦µÄÓÒͼ;¶øË«Ä¿Æ¥ÅäÀú³ÌÓÉË«Ä¿Æ¥ÅäÍøÂçÍê³É£¬£¬£¬£¬£¬£¬ÎüÊÕ×óͼÒÔ¼°ºÏ³ÉµÄÓÒͼ£¬£¬£¬£¬£¬£¬Õ¹Íû³ö×óͼÿһ¸öÏñËصÄÊÓ²îÖµ,ÏêϸµÄÍøÂç½á¹¹(Èçͼ2Ëùʾ)¡£¡£¡£¡£¡£¡£¡£

ͼ2£º

Ëã·¨ÍøÂç½á¹¹Í¼

Éϰ벿·ÖÍøÂç¶ÔÓ¦ÊÓͼºÏ³ÉÍøÂç

Ï°벿·ÖÍøÂç¶ÔӦ˫ĿƥÅäÍøÂç

ÊÓͼºÏ³ÉÍøÂ磺

Ò»Ñùƽ³£ÇéÐÎÏ£¬£¬£¬£¬£¬£¬×óͼÖеÄÏñËؾù¿ÉÒÔÔÚÓÒͼÖÐÕÒµ½Æ¥ÅäµÄÏñËØ£¬£¬£¬£¬£¬£¬Òò´Ë¿ÉÒÔÊ×ÏÈ°Ñ×óͼƽÒƶà¸ö²î±ðÏñËؾàÀ룬£¬£¬£¬£¬£¬»ñµÃ¶àÕŲî±ðµÄͼƬ£¬£¬£¬£¬£¬£¬ÔÙʹÓÃÉñ¾­ÍøÂçÕ¹Íû×éºÏϵÊý£¬£¬£¬£¬£¬£¬°Ñ¶àÕÅƽÒƺóµÄ×óͼºÍÕ¹ÍûµÄϵÊý×éºÏ»ñµÃÕ¹ÍûµÄÓÒͼ¡£¡£¡£¡£¡£¡£¡£ÏêϸµØ£¬£¬£¬£¬£¬£¬ÊÓͼºÏ³ÉÍøÂç»ùÓÚDeep3D [1] Ä£×Ó£¬£¬£¬£¬£¬£¬Í¼2 ÖеÄÉϰ벿·ÖչʾÁËÊÓͼºÏ³ÉÍøÂçµÄʾÒâͼ¡£¡£¡£¡£¡£¡£¡£ÊäÈëÒ»ÕÅ×óͼ£¬£¬£¬£¬£¬£¬Ê×ÏÈÖ÷¸ÉÍøÂç¶ÔÆäÌáÈ¡²î±ð±ê×¼µÄÌØÕ÷£¬£¬£¬£¬£¬£¬ÔÙ¾­ÓÉÉϲÉÑù²ã°Ñ²î±ð±ê×¼µÄÌØÕ÷ͳһÖÁͳһ¸ö³ß´ç£¬£¬£¬£¬£¬£¬È»ºó¾­ÓÉÀÛ¼Ó²Ù×÷ÈںϳÉÊä³öÌØÕ÷²¢Õ¹Íû³ö¸ÅÂÊÊÓ²îͼ£¬£¬£¬£¬£¬£¬×îºó¾­ÓÉÑ¡ÔñÄ£¿£¿£¿£¿£¿£¿£¿é(selection module)ÍŽá¸ÅÂÊÊÓ²îͼÒÔ¼°ÊäÈëµÄ×óͼ£¬£¬£¬£¬£¬£¬»ñµÃÕ¹ÍûµÄÓÒͼ¡£¡£¡£¡£¡£¡£¡£±¾ÎĽÓÄÉL1 Ëðʧº¯ÊýѵÁ·Õâ¸öÍøÂç¡£¡£¡£¡£¡£¡£¡£

Ë«Ä¿Æ¥ÅäÍøÂ磺

Ë«Ä¿Æ¥ÅäÐèÒª°Ñ×óͼÏñËغÍÓÒͼÖÐÆä¶ÔÓ¦ÏñËؾÙÐÐÆ¥Å䣬£¬£¬£¬£¬£¬ÔÙÓÉÆ¥ÅäµÄÏñËزîËã³ö×óͼÏñËضÔÓ¦µÄÉî¶È£¬£¬£¬£¬£¬£¬¶ø֮ǰµÄµ¥Ä¿Éî¶ÈÔ¤¼ÆÒªÁì¾ù²»¿ÉÏÔʽÒýÈëÀàËƵļ¸ºÎÔ¼Êø¡£¡£¡£¡£¡£¡£¡£ÓÉÓÚÉî¶Èѧϰģ×ÓµÄÒýÈ룬£¬£¬£¬£¬£¬Ë«Ä¿Æ¥ÅäËã·¨µÄÐÔÄܽüÄêÀ´»ñµÃÁ˼«´óµÄÌáÉý¡£¡£¡£¡£¡£¡£¡£±¾ÎĵÄË«Ä¿Æ¥ÅäÍøÂç»ùÓÚDispNetC [2] Ä£×Ó, ¸ÃÄ£×ÓÏÖÔÚÔÚKITTIË«Ä¿Æ¥ÅäÊý¾Ý¼¯ÉÏÄܹ»µÖ´ïÀíÏëµÄ¾«¶È£¬£¬£¬£¬£¬£¬ÆäÍøÂçÈçͼ2µÄÏ°벿·ÖËùʾ£¬£¬£¬£¬£¬£¬×óͼÒÔ¼°ºÏ³ÉµÄÓÒͼ¾­Óɼ¸¸ö¾í»ý²ãÖ®ºó£¬£¬£¬£¬£¬£¬»ñµÃµÄÌØÕ÷»á¾­ÓÉ1DÏà¹Ø²Ù×÷(correlation)¡£¡£¡£¡£¡£¡£¡£Ïà¹Ø²Ù×÷±»Ö¤ÊµÔÚË«Ä¿Æ¥ÅäÉî¶ÈѧϰËã·¨ÖÐÆðÒªº¦ÐÔµÄ×÷Ó㬣¬£¬£¬£¬£¬»ùÓÚÏà¹Ø²Ù×÷£¬£¬£¬£¬£¬£¬±¾ÎÄÒªÁìÏÔʽµØÒýÈ뼸ºÎÔ¼Êø;Æä»ñµÃµÄÌØÕ÷ͼºÍ×óͼÌáÈ¡µ½µÄÌØÕ÷ͼ¾ÙÐÐÆ´½Ó×÷Ϊ±àÂë-½âÂëÍøÂç(encoder-decoder network)µÄÊäÈ룬£¬£¬£¬£¬£¬²¢×îÖÕÕ¹ÍûÊÓ²îͼ¡£¡£¡£¡£¡£¡£¡£¸ÃÍøÂçµÄѵÁ·Ò²Í¬ÑùʹÓÃL1Ëðʧº¯Êý¡£¡£¡£¡£¡£¡£¡£

ʵÑéЧ¹û

±¾ÎÄÔÚKITTI¹ûÕæÊý¾Ý¼¯É϶ÔÌá³öµÄÄ£×Ó¾ÙÐÐÑéÖ¤£¬£¬£¬£¬£¬£¬×ñÕÕEigenµÈÈË[3]µÄʵÑéÉèÖ㬣¬£¬£¬£¬£¬°Ñ697ÕÅͼƬ×÷Ϊ²âÊÔͼƬ£¬£¬£¬£¬£¬£¬ÆäÓàµÄÊý¾Ý×÷ΪѵÁ·Í¼Æ¬£¬£¬£¬£¬£¬£¬´Ó¶¨Á¿ºÍ¶¨ÐÔÁ½·½ÃæÁÙËùÌá³öµÄÄ£×Ó¾ÙÐÐÑéÖ¤¡£¡£¡£¡£¡£¡£¡£

ÊýֵЧ¹û

±í1×ܽáÁ˱¾ÎÄÄ£×ÓºÍÆäËûÏÖÓÐÒªÁìЧ¹ûµÄ±ÈÕÕ£¬£¬£¬£¬£¬£¬¿ÉÒÔ¿´³ö£¬£¬£¬£¬£¬£¬±¾ÎÄÄ£×ÓÔÚ´ó´ó¶¼Ö¸±êÉϾùµÖ´ïÌìÏÂÁìÏÈˮƽ¡£¡£¡£¡£¡£¡£¡£ÆäÖУ¬£¬£¬£¬£¬£¬¾ÍARDÖ¸±êÀ´Ëµ£¬£¬£¬£¬£¬£¬Ìá³öµÄÄ£×Ó±È֮ǰ×îºÃµÄÒªÁìÎó²î¼õС16.8%(0.094 vs. 0.113);±íÖÐͬʱҲÏÔʾ£¬£¬£¬£¬£¬£¬¾­Óɶ˵½¶ËÓÅ»¯Ö®ºó£¬£¬£¬£¬£¬£¬SVSÄ£×ÓµÄÐÔÄÜÄܹ»½øÒ»²½»ñµÃÌáÉý¡£¡£¡£¡£¡£¡£¡£

±í1£º

KITTIÊý¾Ý¼¯ÉÏSVSÄ£×ÓºÍÆäËûÒªÁìµÄÊýֵЧ¹û

±íÖмӴÖÌåÏÖÐÔÄÜ×îºÃµÄЧ¹û

¿ÉÊÓ»¯Ð§¹û±ÈÕÕ

ͼ3ÏÔʾÁËÔÚKITTI Eigen²âÊÔ¼¯ÉϵÄÉî¶ÈÔ¤¼ÆЧ¹ûµÄ¿ÉÊÓ»¯Ð§¹û£¬£¬£¬£¬£¬£¬´ÓͼÖпÉÒÔ¿´³ö±¾ÎÄÌá³öµÄSVSÄ£×ÓÄܹ»»ñµÃÔ½·¢¾«×¼µÄÉî¶Èͼ¡£¡£¡£¡£¡£¡£¡£

ͼ3£º

ÔÚKITTI Eigen²âÊÔ¼¯ÉϵÄÉî¶ÈÔ¤¼ÆЧ¹ûµÄ¿ÉÊÓ»¯

Ìá³öµÄSVSÄ£×ÓÄܹ»»ñµÃÔ½·¢×¼È·µÄÉî¶Èͼ

ÔÚÆäËûÊý¾Ý¼¯ÉÏЧ¹ûµÄ¿ÉÊÓ»¯

ΪÁËÑéÖ¤SVSÄ£×ÓÔÚÆäËûÊý¾Ý¼¯Éϵķº»¯ÄÜÁ¦£¬£¬£¬£¬£¬£¬±¾ÎĽ«ÔÚKITTIÊý¾Ý¼¯ÉÏѵÁ·ºÃµÄSVSÄ£×ÓÖ±½ÓÓ¦ÓÃÖÁCityscapeºÍ Make3DÊý¾Ý¼¯ÉÏ£¬£¬£¬£¬£¬£¬Ð§¹û¿ÉÊÓ»¯Ð§¹û»®·ÖչʾÔÚͼ4¼°Í¼5ÖС£¡£¡£¡£¡£¡£¡£¿£¿£¿£¿£¿£¿£¿ÉÒÔ¿´µ½×ÝÈ»ÔÚѵÁ·Êý¾Ý¼¯ÖÐûÓзºÆð¹ýµÄ³¡¾°£¬£¬£¬£¬£¬£¬±¾ÎÄÒªÁìÈÔÈ»¿ÉÒÔ»ñµÃºÏÀí׼ȷµÄÉî¶ÈÔ¤¼ÆЧ¹û£¬£¬£¬£¬£¬£¬Ö¤ÊµÁ˱¾ÎÄÒªÁì½ÏΪǿʢµÄ·º»¯ÄÜÁ¦¡£¡£¡£¡£¡£¡£¡£

ͼ4£º

ÔÚCityscapeÊý¾Ý¼¯ÉÏÉî¶ÈÔ¤¼ÆЧ¹ûµÄ¿ÉÊÓ»¯

SVSÄ£×ÓÄܹ»ÌìÐÄÀíÏëµÄÉî¶Èͼ

ͼ5£º

ÔÚMake3DÊý¾Ý¼¯ÉÏÉî¶ÈÔ¤¼ÆЧ¹ûµÄ¿ÉÊÓ»¯

±¾ÎÄÌá³öµÄSVSÄ£×Ó¿ÉÒÔ»ñµÃ½ÏΪ׼ȷµÄЧ¹û

ÓëË«Ä¿Æ¥ÅäËã·¨Block-MatchingµÄ±ÈÕÕ£º

ΪÁ˽øÒ»²½È·ÈÏÏÖÔÚÐÔÄÜ×îÓÅÒìµÄµ¥Ä¿Éî¶ÈÔ¤¼ÆÒªÁìºÍË«Ä¿Éî¶ÈÔ¤¼ÆÒªÁìµÄ²î±ð£¬£¬£¬£¬£¬£¬±¾ÎÄÔÚKITTI 2015Ë«Ä¿Æ¥Åä²âÊÔ¼¯ÉϱÈÕÕÁËSVSÄ£×ÓÓëÏÖÓÐ×îÓÅÐÔÄܵĵ¥Ä¿Éî¶ÈÔ¤¼ÆÒªÁìÒÔ¼°Ë«Ä¿Æ¥ÅäBlock-MatchingÒªÁì (OCV-BM)£¬£¬£¬£¬£¬£¬Ïà¹ØЧ¹û×ܽáÔÚ±í2ÖУ¬£¬£¬£¬£¬£¬±¾ÎĵÄSVSÄ£×ÓÊ×´ÎÓâÔ½ÁËË«Ä¿Æ¥ÅäBlock-MatchingËã·¨¡£¡£¡£¡£¡£¡£¡£

±í2£º

ÔÚKITTI 2015Ë«Ä¿Æ¥Åä²âÊÔ¼¯ÉϵÄÊýֵЧ¹û

ÊÖÒÕDZÔÚÓ¦ÓÃ

µ¥Ä¿Éî¶ÈÔ¤¼Æ±ÈÕÕË«Ä¿Éî¶ÈÔ¤¼Æ¾ßÓÐÀû±ã°²ÅÅ¡¢±¾Ç®µÍµÈÓŵ㣬£¬£¬£¬£¬£¬ÔÚÐí¶àÁìÓòÓÐן»ºñµÄDZÔÚÓ¦Óó¡¾°£¬£¬£¬£¬£¬£¬ÈçÈýάÖØÐÞ¡¢ÔöÇ¿ÏÖʵµÈ¡£¡£¡£¡£¡£¡£¡£

a) ÈýάÖØÐÞ

b) ÔöÇ¿ÏÖʵ

½áÂÛ

±¾ÎÄÌá³öÒ»ÖÖ¼òÆÓ¶øÓÐÓõĵ¥Ä¿Éî¶ÈÔ¤¼ÆÄ£×Ó¡ª¡ªµ¥ÊÓͼ˫ĿƥÅä(SVS)¡£¡£¡£¡£¡£¡£¡£¸ÃÄ£×Óͨ¹ý°Ñµ¥Ä¿Éî¶ÈÔ¤¼ÆÎÊÌâÆÊÎöΪÁ½¸ö×ÓÎÊÌ⣬£¬£¬£¬£¬£¬¼´ÊÓͼºÏ³ÉÎÊÌâºÍË«Ä¿Æ¥ÅäÎÊÌ⣬£¬£¬£¬£¬£¬×èÖ¹°ÑÉñ¾­ÍøÂçÄ£×ÓÖ±½Ó×÷ΪºÚºÐʹÓ㬣¬£¬£¬£¬£¬Ìá¸ßÁËÄ£×ӵĿÉÚ¹ÊÍÐÔ¡£¡£¡£¡£¡£¡£¡£Í¬Ê±£¬£¬£¬£¬£¬£¬ÎªÁ˸üºÃµÄ½â¾öÕâÁ½¸ö×ÓÎÊÌ⣬£¬£¬£¬£¬£¬ÏÔʽµØ°Ñ¼¸ºÎ±ä»»±àÂëµ½Á½¸ö×ÓÍøÂçÖУ¬£¬£¬£¬£¬£¬ÌáÉýÍøÂçÄ£×ӵıí´ïÄÜÁ¦¡£¡£¡£¡£¡£¡£¡£ÊµÑéЧ¹ûÅú×¢£¬£¬£¬£¬£¬£¬¸ÃÒªÁì½öʹÓÃÉÙÁ¿´øÉî¶È±êÇ©µÄѵÁ·Êý¾Ý£¬£¬£¬£¬£¬£¬¾ÍÄܹ»ÓâÔ½ËùÓÐ֮ǰµÄµ¥Ä¿Éî¶ÈÔ¤¼ÆÒªÁ죬£¬£¬£¬£¬£¬²¢ÇÒÊ״νöʹÓõ¥Ä¿Êý¾Ý¾ÍÁè¼ÝË«Ä¿Æ¥ÅäËã·¨Block-MatchingµÄÐÔÄÜ£¬£¬£¬£¬£¬£¬ÔÚÖÚ¶àÁìÓòÖÐÓÐן»ºñµÄDZÔÚÓ¦Óᣡ£¡£¡£¡£¡£¡£

²Î¿¼ÎÄÏ×

[1] J. Xie, R. Girshick, and A. Farhadi. Deep3d: Fully automatic 2d-to-3d video conversion with deep convolutional neural networks. In ECCV, 2016

[2] N. Mayer, E. Ilg, P. Hausser, P. Fischer, D. Cremers, A. Dosovitskiy, and T. Brox. A large dataset to train convolutional networks for disparity, optical flow, and scene flow estimation. In CVPR, 2016.

[3] D. Eigen, C. Puhrsch, and R. Fergus. Depth map prediction from a single image using a multi-scale deep network. In NIPS, 2014.